CN101690011B - Improve reliability and reduce latency in wireless networks - Google Patents

Abstract

A mesh communication network for use in, for example, a process control plant, includes a plurality of network devices that transmit and receive data according to a network schedule defined with a set of concurrent overlapping superframes and along a set of graphs that define communication paths between pairs of network devices. A network manager, residing either within or outside of a communication network, develops a routing scheme for the network by analyzing the topology of the network and defining a set of graphs for routing or transmitting data between nodes of the network, where each graph includes one or more communication paths between pairs of network devices. Concurrently or subsequently, the network manager defines a network schedule in view of at least transmission requirements, power availability, and signal quality at the various network devices. If desired, the network manager may begin defining the network schedule upon completion of the definition of the graph of the communication network, such that the network manager may define the network schedule in view of the defined graph and the transmit power, etc., parameters associated with the various network devices.

Description

Improve reliability and reduce latency in wireless networks

technical field

This patent relates generally to wireless communications and, more particularly, to improving reliability in wireless networks.

Background technique

Communication protocols rely on various routing techniques to transport data between communication endpoints on a communication network. Communication or network protocols and corresponding routing strategies are usually selected in view of a variety of factors, such as knowledge of the network topology, size of the network, type of medium used as a signal carrier, security and reliability requirements, tolerable transmission delays, and The type of devices forming the network. Due to a number of these factors, typical routing techniques satisfy some design goals at the expense of others. For example, certain routing techniques may provide a high level of reliability in data delivery, but may also require relatively high overhead. Thus, despite the many known methods for routing and the many protocols compatible with these routing methods, there still exist communication networks with specific requirements which are not fully satisfied by any available routing methods and protocols. Moreover, as new types of communication networks with increasing demands for efficiency, throughput, and reliability emerge in a variety of industrial and commercial applications, designers and developers often encounter problems that cannot be achieved with existing protocols and associated routing techniques. New problems that are easily solved.

In general, a communication network includes nodes that are senders and receivers of data sent over communication paths (hardwired or wireless communication paths) connecting the nodes. In addition, communication networks typically include dedicated routers responsible for directing traffic between nodes, and optionally dedicated devices responsible for configuring and managing the network. Some or all of these nodes may also be adapted to act as routers, directing traffic sent between other network devices. Network devices may be wired or wirelessly connected to each other, and network devices may have different routing and transmission capabilities than some nodes within the network. For example, a dedicated router may be capable of high volume transmissions, while certain nodes may only be able to transmit and receive relatively little traffic during the same period of time. Additionally, connections between nodes on the network may have different throughput capabilities and different attenuation characteristics. Fiber optic cables, for example, may be able to be used to provide orders of magnitude higher bandwidth than wireless links due to differences in the inherent physical limitations of the medium.

In order for a node to send data to another node on a typical network, the complete path from source to destination, or the directly relevant portion of that path, must be known. For example, the World Wide Web (WWW) allows pairs of computer hosts to communicate over long distances without either host knowing the full path before sending the information. Instead, hosts are configured with information about their assigned gateways and dedicated routers. Specifically, the Internet Protocol (IP) provides network layer connectivity to the WWW. IP defines a sub-protocol called Address Resolution Protocol (ARP), which provides a local table at each host specifying routing rules. Thus, a typical host connected to the WWW or similar wide area network (WAN) would know to route all packets with a predetermined address matching a pre-configured pattern to host A and the rest to host B. Similarly, the direct hosts or "hops" that forward these packets also perform part of the routing decisions and generally direct the data in the general direction of the destination.

In most network protocols, most or all network devices are assigned unique addresses sufficiently to enable hosts to exchange information in an unambiguous manner. At least in the case of unicast (one-to-one) delivery, the destination address must be specified at the source. For this reason, network protocols usually define strict addressing schemes. As one of ordinary skill in the art will recognize, modifying or extending an addressing scheme is a complex and expensive process. For example, the transition from version 4 (IPv4) to version 6 (IPv6) of the IP protocol required a major update of much of the infrastructure supporting IPv4. On the other hand, defining an addressing scheme with a large capacity for a small network creates unnecessary overhead. Therefore, a network protocol ideally adapted to a particular application provides a sufficient number of possible addresses without excessive overhead in data transfer.

协议。一般而言，

协议支持专用导线或导线组上的数字和模拟混合信号，在专用导线或导线组上，在线过程信号(如控制信号、传感器测量结果等)作为模拟电流信号(例如范围从4到20毫安)被提供，并且诸如设备数据、对设备数据的请求、配置数据、报警以及事件数据等等之类的其它信号，作为叠加或复用到与模拟信号相同的导线或导线组上的数据信号被提供。然而，HART协议当前需要使用专用、硬线通信线路，这导致需要在加工厂内大量布线。In short, there are a large number of factors that affect the implementation of a particular protocol in a particular industry. In the process control industry, it is known to use standard communication protocols to enable devices made by different manufacturers to communicate with each other in an easy-to-use and implement manner. A well-known communication standard used in the process control industry is the High Speed Addressable Remote Transducer (HART) Communications Base Protocol, commonly referred to as

protocol. Generally speaking,

Protocol supports mixed digital and analog signals on dedicated wires or wire sets on which in-line process signals (e.g. control signals, sensor measurements, etc.) as analog current signals (e.g. range from 4 to 20 mA) and other signals such as device data, requests for device data, configuration data, alarm and event data, etc., are provided as data signals superimposed or multiplexed onto the same wire or groups of wires as the analog signals . However, the HART protocol currently requires the use of dedicated, hard-wired communication lines, resulting in the need for extensive wiring within the fab.

Over the past few years, there has been a growing trend to incorporate wireless technology into various industries including, in some limited ways, the process control industry. However, within the process control industry, there are formidable barriers that limit the full integration, acceptance, and use of wireless technologies. Specifically, the process control industry requires a very reliable process control network, because the loss of a signal can lead to a loss of control of the plant, resulting in catastrophic consequences, including explosions, the release of deadly chemicals or gases, and so on. For example, US Patent No. 6,236,334 to Tapperson et al. discloses the use of wireless communication as a secondary or backup communication path or for sending non-critical or redundant communication signals in the process control industry. Also, many advances have been made in the use of wireless communication systems that are generally applicable to the process control industry, but have not yet been developed to allow or provide a reliable and in some cases completely wireless communication network within a process plant. Applied in the process control industry. For example, US Patent Application Publication Nos. 2005/0213612, 2006/0029060, and 2006/0029061 disclose various aspects of wireless communication technologies related to general wireless communication systems.

Similar to wired communication, wireless communication protocols are expected to provide an efficient, reliable and secure method of information exchange. Of course, most of the methods developed to solve these important problems in wired networks are not suitable for wireless communication due to the shared and open nature of the medium. Further, besides the typical objective facts behind wired communication protocols, wireless protocols also face other requirements regarding the problems of interference and coexistence of several networks using the same part of the radio spectrum. Also, some wireless networks operate in unlicensed or public frequency bands. Therefore, protocols serving such networks must be able to detect or resolve issues related to frequency (channel) contention, radio resource sharing and negotiation, and the like.

In the process control industry, developers of wireless communication protocols face additional challenges, such as achieving backward compatibility with wired devices, supporting previous wired versions of the protocol, providing conversion services for devices retrofitted with wireless communicators, and providing Routing technology that guarantees reliability and availability at the same time. At the same time, there are still quite a few process control applications in which there is little if any in-situ measurement. Currently, these applications rely on observed measurements (eg, water levels are rising) or inspections (eg, periodic maintenance of air conditioning units, pumps, fans, etc.) to spot anomalies. In order to take action, operators often need face-to-face discussions. The application of many of these applications could be greatly simplified if measurement and control equipment were used; however, current measurement equipment often requires power, communication infrastructure, configuration, and support infrastructure that is simply not available.

According to yet another aspect, the process control industry requires that a communication protocol serving a particular process control network be able to accommodate field devices having different data transmission requirements, priorities, and power capabilities. In particular, some process control systems may include measurement devices that report measurements frequently (eg, several times per second) to a centralized controller or to another field device. Meanwhile, another device in the same system may only report measurements, alarms or other data every hour. However, both devices may require that individual measurement reports be propagated to a destination host, such as a controller, workstation or equivalent field device, with as little time and bandwidth overhead as possible.

Contents of the invention

A mesh communication network used, for example, in a process control plant includes a plurality of network devices according to a network schedule defined by a set of concurrent overlapping superframes and along a set of network devices defining communication paths between pairs of network devices. Figure to send and receive data. In some embodiments, a network manager residing inside or outside of a communication network develops a network manager for the network by analyzing the topology of the network and defining a set of graphs for routing or sending data between the nodes of the network. Routing schemes, where each graph includes one or more communication paths between pairs of network devices. Concurrently or subsequently, the network manager defines a network schedule in view of at least transmission requirements, power availability, and signal quality at the various network devices. If necessary, the network manager can start defining the network schedule table after completing the definition of the graph of the communication network, so that the network manager can define the network schedule in view of the defined graph and parameters such as transmit power associated with each network device surface.

In some embodiments, the mesh communication network is a wireless network, the graph is a directed graph, and accordingly, the communication paths are unidirectional communication paths. Network devices of a mesh communication network can originate and route data on behalf of other network devices. In another embodiment, the mesh communication network conforms to a star grid topology, in which some network devices can only receive data or originate data, and some network devices can receive data, originate data, and Relay data between network devices.

According to another aspect, each superframe includes a number of communication slots of predetermined duration, and each superframe repeats immediately after all communication slots in the superframe have occurred. In general, the total number of slots in each superframe defines the length of the superframe, with each particular slot having a relative number of slots corresponding to the number of slots present in the superframe preceding the particular slot. In one embodiment, each time slot includes a transmission period in which one or more network devices transmit data and an acknowledgment period in which one or more network devices transmit acknowledgments corresponding to the transmitted data. In some embodiments, the network device also performs a pure channel evaluation to determine whether a particular common timeslot is available for transmission.

After defining the group graph and communication schedule, the network manager communicates the relevant routing and scheduling information to some or all network devices (nodes), so that packets sent from one network device to another network device can be sent according to The graph is defined by the network manager and is routed appropriately in time slots allocated by the network manager. In another embodiment, the functions of analyzing the network and obtaining the network topology are distributed among at least several network devices, so that one or more network devices participate in defining a one-way or two-way graph.

The length of each superframe may correspond to the transmission requirements of a particular network device. In this case, the dedicated service defines superframes according to the needs of the network devices and external hosts communicating with the network devices, and allocates time slots within each of these superframes. In one embodiment, a dedicated service associates a network device with one or more time slots of a particular superframe such that the network device can send or receive data during each occurrence of that time slot. A network device may participate in multiple superframes, if necessary, to send data intended for that network device and to forward data between other network devices.

If desired, dedicated services can dynamically create and destroy superframes in view of changes in network conditions, such as data bursts, congestion, block transfers, and network devices entering or leaving the network. Additionally, a network device or dedicated service can effectively invalidate a superframe without destroying the superframe by issuing specific commands. If desired, the dedicated service may be a software entity running on a dedicated physical host, or the dedicated service may run on a gateway device that connects the wireless mesh network to external networks or hosts.

A network schedule may include multiple communication channels, and in some cases, each communication channel may correspond to a unique carrier radio frequency. Each network device may have a single schedule including relative slot numbers and communication channel identifiers, and the single schedule may specify which network device to use to transmit process data, route data originating from another network device, receive device-specific A separately scheduled time slot for receiving broadcast data or data. In some embodiments, a single schedule for a network device may specify timeslots associated with several different communication channels during different superframe cycles such that a network device has the same relative number of slots for a particular superframe Send or receive data on different communication channels in different time slots. In some of these embodiments, the network device iterates in a predefined manner among several channels associated with a particular time slot according to a corresponding single schedule. In other embodiments, the network schedule does not assign different communication channels to the same time slot.

If desired, a dedicated service can create long superframes for transient devices that wake up periodically according to a predefined schedule, thereby eliminating the need for the transient device to negotiate resources for each transmit session. In yet another embodiment, a transient device conserves power by only sending data according to the necessary update rate for the transient device.

In yet another embodiment, at least some of the network devices are field devices that perform measurement or control functions in a process control environment. Each of these field devices is specified a specific update rate, or frequency at which process data is communicated to another network device. In this case, a dedicated service defines a superframe according to the update rate of the field devices. In addition, field devices can negotiate temporary changes in time slot allocation, if desired, due to the occurrence of transient conditions requiring higher than normal or lower than normal bandwidth.

In one embodiment, the management entity responsible in part for defining the unidirectional graph is a dedicated network manager and may be implemented as a software module running on a host residing inside or outside the network. In another embodiment, the network manager may be a dedicated physical device communicatively connected to the network. In yet another embodiment, the network manager may be distributed among several devices residing inside or outside the network.

Description of drawings

Figure 1 schematically illustrates a wireless network connected to a factory automation network through a gateway device, providing communication between field devices and a router device, and using the routing techniques of the present disclosure.

FIG. 2 is a schematic diagram of the layers of the WirelessHART protocol that may be used in the wireless network shown in FIG. 1 .

Figure 3 is a block diagram illustrating segments of a communication slot defined according to one of the embodiments discussed herein.

4 is a block diagram illustrating an exemplary relationship of time slots of a three-slot superframe to a number of communication devices.

Fig. 5 schematically shows the relationship of time slots of an exemplary superframe to several communication channels.

Fig. 6 schematically shows a block diagram of an exemplary superframe definition comprising several concurrent superframes of different lengths.

Fig. 7 schematically shows a star network topology to which a wireless network such as the network shown in Fig. 1 or Fig. 3 may follow.

Fig. 8 schematically shows a mesh network topology to which a wireless network such as the network shown in Fig. 1 or Fig. 3 may follow.

Fig. 9 schematically shows a star mesh network topology to which a wireless network such as the network shown in Fig. 1 or Fig. 3 may follow.

10 is a block diagram illustrating redundant upstream data paths in a wireless network utilizing certain graph routing techniques of the present disclosure.

11 is a block diagram illustrating redundant downstream data paths in a wireless network utilizing certain graph routing techniques of the present disclosure.

12 is a block diagram illustrating redundant upstream and downstream data paths in a wireless network utilizing certain graph routing techniques of the present disclosure.

Figure 13 schematically illustrates a wireless network with several network devices communicating data according to a device-specific schedule and utilizing certain routing and scheduling techniques of the present disclosure.

FIG. 14 shows an example superframe definition for the two devices shown in FIG. 13 .

Figures 15 and 16 show the time slot allocation for several types of data that may be transmitted in the management superframe of the network shown in Figure 13 .

17-20 illustrate example communication schedule definitions for each device shown in FIG. 13 .

Detailed ways

Figure 1 illustrates an exemplary network 10 in which the scheduling and routing techniques described herein may be used. Specifically, the network 10 may include a factory automation network 12 connected to a wireless communication network 14 . The factory automation network 12 may include one or more fixed workstations 16 and one or more portable workstations 18 connected by a communication backbone 20, which may utilize Ethernet, RS-485, Profibus DP, or utilize suitable communication hardware and agreement to achieve. These workstations and other devices forming the factory automation network 12 may provide factory personnel with various control and supervisory functions, including access to devices in the wireless network 14 . Factory automation network 12 and wireless network 14 may be connected via gateway device 22 . More specifically, gateway device 22 may be wired to backbone 20 and may communicate with factory automation network 12 using any suitable (eg, known) communication protocol. Can be used in any other desired manner (e.g., as a stand-alone device, a card that can be plugged into an expansion slot of a host workstation 16 or 18, as part of an input/output (I/O) subsystem of a PLC-based or DCS-based system, etc. Gateway device 22 , implemented as a wireless network 14 , may provide applications that run on network 12 and have access to various devices of wireless network 14 . In addition to protocol and command conversion, gateway device 22 may also provide time slots and superframes (communications evenly spaced in time) of a scheduling scheme associated with the wireless protocol implemented in network 14 (referred to herein as the WirelessHART protocol). time slot group) used for synchronous timing.

In some configurations, the network 10 may include more than one gateway device 22 to increase the efficiency and reliability of the network 10 . Specifically, multiple gateway devices 22 may provide additional bandwidth for communications between wireless network 14 and factory automation network 12 and the outside world. Alternatively, gateway device 22 may request bandwidth from an appropriate network service as needed for gateway communications within wireless network 14 . A network manager software module 27, which may reside in the gateway device 22, may further re-evaluate the required bandwidth while the system is running. For example, gateway device 22 may receive requests from hosts residing outside wireless network 14 to retrieve large amounts of data. Gateway device 22 may then request network manager 27 to allocate additional bandwidth to accommodate the transaction. For example, gateway device 22 may issue an appropriate service request. Gateway device 22 may then request network manager 27 to release the bandwidth when the transaction is complete.

With continued reference to FIG. 1 , the wireless network 14 may include one or more field devices 30 - 36 . Typically, process control systems such as those used in chemical, petroleum, or other process plants include field devices such as valves, valve positioners, switches, sensors (eg, temperature, pressure, and flow rate sensors) pumps, fans, and the like. Generally, field devices perform physical control functions within the process, such as opening or closing valves or obtaining measurements of process parameters. In the wireless communication network 14, the field devices 30-36 are producers and consumers of wireless communication packets.

Devices 30-36 may communicate using a wireless communication protocol that provides similar functionality of a wired network, with similar or improved operational performance. Specifically, the protocol may enable the system to perform process data monitoring, critical data monitoring (with more stringent performance requirements), calibration, equipment status and diagnostic monitoring, field equipment troubleshooting, commissioning, and supervisory process control. However, applications that perform these functions typically require wireless network 14 supported protocols to provide fast updates when necessary, move large amounts of data when needed, and support network devices that join wireless network 14, even if only temporarily for commissioning and repairs. care work.

Fieldbus、PROFIBUS、DeviceNet等等。在这些实施例中，WHA 50支持在协议栈的较下层上的协议转换。另外，假设单个WHA 50也可以充当多路复用器并且可以支持多个HART或非HART设备。Network 14 may include non-wireless devices, if desired. For example, field device 38 of FIG. 1 may be a legacy 4-20 mA device, and field device 40 may be a conventional wired HART device. To communicate within network 14 , field devices 38 and 40 may connect to WirelessHART network 14 via WirelessHART adapter (WHA) 50 or 50A. Additionally, WHA 50 may support other communication protocols, for example,

Fieldbus, PROFIBUS, DeviceNet, etc. In these embodiments, the WHA 50 supports protocol translation at the lower layers of the protocol stack. Additionally, assume that a single WHA 50 can also act as a multiplexer and can support multiple HART or non-HART devices.

In general, network manager 27 may be responsible for adapting wireless network 14 to changing conditions and for scheduling communication resources. As network devices join and leave the network, the network manager 27 can update its internal model of the wireless network 14 and use this information to generate communication schedules and communication routes. In addition, network manager 27 may consider the overall performance of wireless network 14 as well as diagnostic information to adapt wireless network 14 to changes in topology and communication requirements. Once the network manager 27 has generated the overall communication schedule, all or individual parts of the overall communication schedule may be transmitted from the network manager 27 to the network devices by a series of commands.

To further increase bandwidth and improve reliability, the gateway device 22 can be functionally divided into a virtual gateway 24 and one or more network access points 25, which can be separate physical devices that communicate with the gateway device 22 by wire . However, while FIG. 1 shows a wired connection 26 between physically separate gateway network 22 and access point 25, it will be appreciated that elements 22-26 may also be provided as an integral device. Because network access point 25 may be physically separate from gateway device 22 , access point 25 may be strategically placed at several different locations on network 14 . In addition to increasing bandwidth, multiple access points 25 can increase the overall reliability of network 14 by utilizing other access points 25 to compensate for possible poor signal quality at one access point 25 . Having multiple access points 25 also provides redundancy in the event that one or more access points 25 fail.

In addition to allocating bandwidth, or otherwise bridging networks 12 and 14 , gateway device 22 may also perform one or more management functions in wireless network 14 . As shown in FIG. 1 , the network manager software module 27 and the security manager software module 28 may be stored in and executed in the gateway device 22 . Alternatively, network manager 27 and/or security manager 28 may run on one of hosts 16 or 18 in factory automation network 12 . For example, network manager 27 may run on host 16 and security manager 28 may run on host 18 . Network manager 27 may be responsible for configuring network 14, scheduling communications between wireless devices, managing routing tables associated with those wireless devices, monitoring the overall health of wireless network 14, reporting the health of wireless network 14 to workstations 16 and 18, and Other administrative and supervisory functions. While it may be sufficient to have a single network manager 27 operating in the network 14, redundant network managers 27 may similarly be provided to protect the wireless network from unexpected equipment failures. At the same time, security manager 28 may be responsible for protecting wireless network 14 from malicious or inadvertent intrusion by unauthorized devices. To this end, security manager 28 may manage authentication codes, verify authorization information provided by devices attempting to join wireless network 14, update transient security data such as expired keys, and perform other security functions.

With continued reference to FIG. 1 , the wireless network 14 may include one or more field devices 30 - 36 . Typically, process control systems such as those used in chemical, petroleum, or other process plants include field devices such as valves, valve positioners, switches, sensors (eg, temperature, pressure, and flow rate sensors), pumps, fans, and the like. Field devices perform physical control functions within the process, such as opening or closing valves or taking measurements of process parameters. In the wireless communication network 14, the field devices 30-36 are producers and consumers of wireless communication packets.

Devices 30-36 may communicate using a wireless communication protocol that provides similar functionality of a wired network, with similar or improved operational performance. Specifically, the protocol may enable the system to perform process data monitoring, critical data monitoring (with more stringent performance requirements), calibration, equipment status and diagnostic monitoring, field equipment troubleshooting, commissioning, and supervisory process control. However, the applications that perform these functions typically require that the protocols supported by the wireless network 14 provide fast updates when necessary, move large amounts of data when needed, and support network devices that join the wireless network 14, even if only temporarily for commissioning and maintenance work.

In one embodiment, the wireless protocol supporting the network devices 30-36 of the wireless network 14 is an extension of the known wired HART protocol, a widely accepted industry standard that maintains simple workflows and practices for wired environments. In this sense, network devices 30-36 may be considered to be WirelessHART devices. The same tools used for wired HART devices can be easily adapted for wireless devices 30-36 by simply adding a new device description file. In this way, the wireless protocol can minimize training and simplify maintenance and support by leveraging experience and knowledge gained using the wired HART protocol. In general, it can be convenient to adapt the protocol to wireless applications so that most applications running on the device do not "notice" the transition from a wired network to a wireless network. Clearly, such transparency greatly reduces the cost of upgrading networks, and more generally, the costs associated with developing and supporting equipment that can be used with such networks. Some additional benefits of wireless extensions to the well-known HART protocol include: access to measurements that are difficult or expensive to obtain with wired devices and the ability to install on laptops, handhelds, workstations, etc. The system software configuration and the ability to operate the instrument. Another benefit is the ability to route diagnostic alerts from the wireless device back through the communications infrastructure to a centrally located diagnostic center. For example, each heat exchanger in a process plant might be installed with a WirelessHART device and when the exchanger detects a problem, the end user and supplier could be alerted. An added benefit is the ability to monitor conditions that present serious health and safety concerns. For example, a WirelessHART device might be placed on a road in a flood zone and used to alert authorities or drivers about water levels. Other benefits include: access to a wide range of diagnostic warnings, and the ability to store trended and calculated values at the WirelessHART device so that they can be transmitted to the host computer when communication with the device is established. In this way, the WirelessHART protocol can provide a platform that enables host applications to wirelessly access existing HART-enabled field devices, and the WirelessHART protocol can support battery-operated, wireless-only HART-enabled field devices. deploy. The WirelessHART protocol can be used to establish a wireless communication standard for process applications and can further expand the applications of HART communication and the benefits the protocol provides to the process control industry by enhancing basic HART technology to support wireless process automation applications.

Referring again to FIG. 1, the field devices 30-36 may be WirelessHART field devices, each field device 30-36 taking the form of an integral unit and supporting all layers of the WirelessHART protocol stack. For example, in network 14, field device 30 may be a WirelessHART flow meter, field device 32 may be a WirelessHART pressure sensor, field device 34 may be a WirelessHART valve positioner, and field device 36 may be a WirelessHART pressure sensor. Importantly, wireless devices 30-36 can support all of the HART features that users have seen from the wired HART protocol. As will be appreciated by those skilled in the art, one of the core strengths of the HART protocol lies in its strict interoperability requirements. In some embodiments, all WirelessHART devices include core mandatory capabilities to allow equivalent device types (eg, manufactured by different manufacturers) to be interchanged without compromising system operation. Furthermore, the WirelessHART protocol is backward compatible with core HART technologies such as Device Description Language (DDL). In a preferred embodiment, all WirelessHART devices should support DDL, which ensures that end users directly have the tools to start using the WirelessHART protocol.

Fieldbus、PROFIBUS、DeviceNet等等。在这些实施例中，WHA 50支持在协议栈的较下层上的协议转换。另外，假设单个WHA 50也可以充当多路复用器并且可以支持多个HART或非HART设备。Network 14 may include non-wireless devices, if desired. For example, field device 38 of FIG. 1 may be a legacy 4-20 mA device, and field device 40 may be a conventional wired HART device. To communicate within network 14 , field devices 38 and 40 may connect to Wireless HART network 14 via Wireless HART adapter (WHA) 50 . Additionally, WHA 50 may support other communication protocols, for example,

Fieldbus, PROFIBUS, DeviceNet, etc. In these embodiments, the WHA 50 supports protocol translation at the lower layers of the protocol stack. Additionally, assume that a single WHA 50 can also act as a multiplexer and can support multiple HART or non-HART devices.

Plant employees can additionally use handheld devices to install, control, monitor and maintain network devices. In general, a handheld device is a portable device that can connect to the wireless network 14 directly or through a gateway device 22 as a host on the factory automation network 12 . As shown in FIG. 1 , a wireless HART-connected handheld device 55 can communicate directly with the wireless network 14 . When operating with the formed wireless network 14, the handheld device 55 may simply join the network 14 as another WirelessHART field device. When operating with a target network device that is not connected to a WirelessHART network, the handheld device 55 can operate as a combination of gateway device 22 and network manager 27 by forming its own wireless network with the target network device.

A handheld device (not shown) connected to the factory automation network may be used to connect to the factory automation network 12 through known networking techniques such as Wi-Fi. This device communicates with network devices 30-40 through gateway device 22, or workstations 16 and 18 communicate with devices 30-40 in the same manner as an external factory automation server (not shown).

Additionally, wireless network 14 may include router device 60, which is a network device that forwards packets from one network device to another. A network device that is acting as a router device uses an internal routing table to direct routing, that is, to decide which network device a particular packet should be sent to. In those embodiments where all devices on wireless network 14 support routing, a stand-alone router such as router 60 may not be required. However, it may be beneficial to add one or more dedicated routers 60 to the network 14 (eg, to expand the network, or to conserve power for field devices in the network).

All devices directly connected to wireless network 14 may be referred to as network devices. Specifically, wireless field devices 30-36, adapter 50, router 60, gateway device 22, access point 25, and wireless handheld device 55 are network devices for routing and scheduling purposes, each of which forms a wireless A node of the network 14 . In order to provide a very robust and easily scalable wireless network, all devices in the network can support routing and each network device can be globally identified with a substantially unique address, such as a HART address. Network manager 27 may contain a complete list of network devices and may assign each device a short, network-unique 16-bit (for example) alias. Additionally, each network device may store information related to update ("scanning") rates, connection sessions, and device resources. In short, each network device maintains up-to-date information related to routing and scheduling within the wireless network 14 . The network manager 27 may communicate this information to network devices whenever a new device joins the network or whenever the network manager 27 detects or initiates a change in the topology or schedule of the wireless network 14 .

Additionally, each network device may store and maintain a list of neighbor devices that the network device has identified during listening operations. In general, a neighbor of a network device is another network device of any type potentially capable of establishing a connection with the network device according to standards enforced by the respective network. In the case of a WirelessHART network 14, the connection is a direct wireless connection. However, it will be understood that a neighbor device may also be a network device connected to a particular device in a wired manner. As will be discussed later, network devices may facilitate their discovery by other network devices through advertisements or special messages sent out at specified time periods. Network devices operatively connected to wireless network 14 have one or more neighbors, which network devices may select based on the strength of the advertised signal or according to some other criteria.

In the example shown in FIG. 1 , each of a pair of network devices connected by a direct wireless connection 65 recognizes the other as a neighbor. The network devices of the wireless network 14 may form a large number of device-to-device connections 65 . The possibility and desire to establish a direct wireless connection 65 between two network devices is determined by factors such as the physical distance between these nodes, obstacles between these nodes (devices), signal strength at each of these two nodes Wait for a number of factors to determine. Typically each radio connection 65 is characterized by a large set of parameters related to transmission frequency, method of accessing radio resources, etc. Those of ordinary skill in the art will recognize that, in general, wireless communication protocols may operate on designated frequencies, such as those assigned by the Federal Communications Commission (FCC) in the United States, or in unlicensed radio frequency bands (e.g., 2.4GHz). Although the systems and methods discussed herein are applicable to wireless networks operating on any given frequency or range of frequencies, the exemplary embodiments discussed below relate to wireless networks 14 operating on unlicensed or shared portions of the radio spectrum. . According to this embodiment, the wireless network 14 can be easily driven or tuned to operate in specific unlicensed frequency ranges as desired.

With continued reference to FIG. 1 , two or more direct wireless connections 65 may form communication paths between nodes that cannot form direct wireless connections 65 . For example, direct wireless connection 65A between WirelessHART handheld device 55 and WirelessHART device 36 , along with direct wireless connection 65B between WirelessHART device 36 and router 60 , may form the communication path between devices 55 and 60 . As discussed in more detail below, at least some of these communication paths may be directed communication paths (ie, allowing data to be transmitted in one direction only, or between a pair of devices). Meanwhile, WirelessHART device 36 may be directly connected to each of network devices 55, 60, 32, and to network access points 25A and 25B. In general, network devices operating in wireless network 14 may originate packets, relay packets sent by other devices, or perform both types of operations. As used herein, the term "end device" refers to a network device that does not relay packets sent by other devices, while the term "routing device" refers to a network device that relays packets traveling between other network devices. Of course, the routing device can also originate its own data, or in some cases can also be an end device. Thus, one or several end devices and routing devices, together with several direct connections 65, may form part of a mesh network.

Because a process plant may have hundreds or even thousands of field devices, a wireless network 14 operating in the plant may include a large number of nodes, and in many operating situations an even greater number of direct connections 65 between pairs of nodes. As a result, the wireless network 14 may have a complex mesh topology, and some pairs of devices that do not share a direct connection 65 may have to communicate through many intermediate hops that perform communication between those devices. Therefore, it may sometimes be necessary for a data packet to travel along many direct connections 65 after leaving the source device before reaching the destination device, and each direct connection 65 may add a delay to the total transit time of the data packet. Also, some of these intermediary devices may be located at the intersection of many communication paths of the mesh network. Also, these devices may be responsible for relaying a large number of packets originating from many different devices, possibly in addition to originating its own packets. Thus, a relatively busy intermediate device may not immediately forward a transient data packet, but may queue the packet for a relatively long time before sending the packet to the next node on the corresponding communication path. When the data packet finally reaches the destination device, the destination device may reply with an acknowledgment packet, which may also experience a similar delay. During the propagation of the packet to the destination device and the propagation of the corresponding acknowledgment packet from the destination device back to the originating device, the originating node may not know whether the data has successfully reached the destination device. Furthermore, devices may leave the wireless network 14 due to scheduled maintenance and upgrades or due to unexpected failures, thereby changing the topology of the mesh network and disrupting some of these communication paths. Similarly, these devices may join the wireless network 14, thereby adding an additional direct connection 65. These or other changes to the topology of wireless network 14 may significantly affect the transmission of data between pairs of nodes if not handled in an efficient and timely manner.

Importantly, however, the efficiency with which packets are delivered can largely determine the reliability, safety and overall quality of plant operations. For example, a data packet including a measurement indicating an excessive temperature of a reactor should reach another node, such as a handheld device 55 or even a workstation 16, quickly and reliably so that an operator or controller can take appropriate action right away and Work to resolve hazardous situations if necessary. In order to efficiently utilize the available direct wireless connections 65 and fully adapt to frequently changing network topologies, the network connector 27 can maintain a complete network map 68, define routing schemes connecting at least some pairs of network devices 30-50, and The relevant portion of the routing scheme is communicated to each network device participating in the routing scheme.

Specifically, network manager 27 may define a set of directed graphs that include one or more unidirectional communication paths, assign a graph identifier to each defined directed graph, and may assign the relevant portion of each graph definition to This is communicated to each respective network device, which may then update the locally stored connection table 69 for the particular device. As explained in more detail below, the network devices 30-50 may then route the data packets based on the map identifier included in the packet headers or trailers, etc. of the data packets. If desired, each connection table 69 may only store routing information directly related to the corresponding network device, such that the network device does not know the full definition of the directed graph comprising the network device. In other words, the network device may not be able to "see" networks beyond its immediate neighbors, and in this sense, the network device may not know the complete topology of the wireless network 14 . For example, router device 60 shown in FIG. 1 may store connection table 69A, which may only specify routing information related to neighboring network devices 32 , 36 , 50 , and 34 . At the same time, WHA 50A may store a connection table 69B, which in turn may specify routing information related to WHA 50A's neighbors.

In some cases, the network manager 27 may define dual communication paths between pairs of network devices to ensure that in the event that one of the primary communication path's direct connections 65 becomes unavailable, packets can still follow the secondary communication path reach the destination device. However, some of the direct connections 65 may be shared between the primary and secondary paths of a particular pair of network devices. Also, the network manager 27 may in some cases communicate the entire communication path to use to a network device, which may then originate the data packet and include the complete path information in the header and trailer of the data packet . Preferably, network devices use this routing method for data that does not have strict latency requirements. As discussed in detail below, this approach (referred to herein as "source routing") may not offer the same degree of reliability and flexibility, and, in general, may be characterized by longer delivery delays.

Another core requirement of wireless network protocols (and especially of wireless networks operating in unlicensed frequency bands) is to coexist with minimal disruption with other equipment using the same frequency band. Coexistence is generally defined as the ability of one system to perform tasks in a shared environment in which other systems are able to similarly perform their tasks while complying with the same set of criteria or a different (and possibly unknown) set of criteria. One requirement for coexistence in a wireless environment is the ability of the protocol to maintain communication in the presence of interference in the environment. Another requirement is that the protocol should cause as little interference and disturbance as possible to other communication systems.

In other words, the problem of a wireless system coexisting with the surrounding wireless environment generally has two aspects. The first aspect of coexistence is the way in which the system affects other systems. For example, an operator or developer of a particular system may ask what effect a signal sent by one transmitter has on other radio systems operating in close proximity to the particular system. More specifically, the operator may ask whether the transmitter is disrupting the communication of some other wireless device whenever the transmitter is turned on, or whether the transmitter is spending too much time on the broadcast effectively "monopolizing" (hogging) the bandwidth. Ideally, each transmitter should be a "quiet neighbor" unnoticed by other transmitters. Although this ideal property is rarely, if ever, achieved, wireless systems that create a coexistence environment in which other wireless communication systems can operate reasonably well may be referred to as "good neighbours." A second aspect of coexistence of a wireless system is the system's ability to function reasonably well in the presence of other systems or sources of wireless signals. In particular, the robustness of a wireless system can depend on how well the wireless system prevents interference at these receivers, depending on whether these receivers are easily overloaded by close sources of RF energy, depending on whether these receivers How tolerant it is to occasional bit loss, and similar factors. In certain industries, including the process control industry, there are many important potential applications where data loss is often not tolerated. A wireless system capable of providing reliable communication in a noisy or dynamic radio environment may be referred to as a "forgiving neighbor".

Effective coexistence (ie, being a good neighbor and a tolerant neighbor) depends in part on the efficient use of three dimensions of freedom: time, frequency, and distance. Communications are likely to be successful when they occur 1) when the source of the interference (or other communication system) is quiet; 2) occur at a different frequency than the interfering signal; or 3) occur far enough away from the source of the interference . While a single of these factors may be used to provide a communication scheme in a shared portion of the radio spectrum, a combination of two or all three of these factors may provide a high degree of reliability, security and speed.

Still referring to FIG. 1 , network manager 27 or another application or service running on network 14 or 12 may define a master network schedule 67 for wireless communication network 14 in view of the factors discussed above. The master network schedule 67 may specify the allocation of resources, such as time slots and radio frequencies, for the network devices 25 and 30-55. Specifically, master network schedule 67 may specify when each of network devices 25 and 30-55 transmit process data, route data on behalf of other network devices, listen for management data propagated from network manager 27, and Advertisement data is sent to devices on the wireless network 14. In order to allocate radio resources in an efficient manner, the network manager 27 may define and update the master network schedule 67 in view of the topology of the wireless network 14 . More specifically, network manager 27 may allocate available resources to each of the nodes of wireless network 14 (ie, wireless devices 30-36, 50, and 60) based on the direct wireless connections 65 identified at each node. In this sense, the network manager 27 can define and maintain a network schedule 67 in view of the transmission requirements and routing possibilities at each node.

The main network schedule 67 may divide available radio resources into independent communication channels, and further measure transmission and reception opportunities on each channel in units of, for example, Time Division Multiple Access (TDMA) communication slots. Specifically, wireless network 14 may operate within a frequency band that can, in most cases, be safely associated with several distinct carrier frequencies so that communications on one frequency may communicate with other frequencies within that frequency band. Communications on one frequency occur simultaneously. Those of ordinary skill in the art will understand that in typical applications the carrier frequencies (eg, public radio) are sufficiently separated to prevent interference between adjacent carrier frequencies. For example, in the 2.4GHz band, the IEEE assigns frequency 2.455 to channel number 21 and frequency 2.460 to channel number 22, allowing 5KHz between two adjacent segments of the 2.4GHz band. The main network schedule 67 may thus associate each communication channel with a distinct carrier frequency, which may be the center frequency of a particular segment of the frequency band.

Meanwhile, as commonly used in industries using TDMA technology, the term "slot" refers to a specific duration of time into which a larger period of time is divided to provide a controlled sharing method. For example, one second can be divided into 10 equal 100 millisecond slots. While the master network schedule 67 preferably allocates resources in a single fixed-duration time slot, it is possible to change the duration of these time slots as long as each relevant node of the wireless network 14 is properly notified of the change. . Continuing with the example definition of 10 100ms time slots, two devices can exchange data once per second, during the first 100ms period of each second (that is, the first time slot), one device transmits, and during the fourth During 100ms (ie, the fourth time slot), another device transmits, while the remaining time slots are not occupied. Thus, nodes on the wireless network 14 can identify scheduled transmission and reception opportunities by the transmission frequencies and time slots during which corresponding devices can transmit and receive data.

As part of defining an efficient and reliable network schedule 67, the network manager 27 may logically organize time slots into recurring groups or superframes. As used herein, a superframe may be more precisely understood as a series of equal superframe cycles, each superframe cycle corresponding to a logical grouping of several contiguous time slots forming a continuous period of time. The number of slots within a given superframe defines the length of the superframe and determines how often each slot repeats. In other words, the length of the superframe multiplied by the duration of a single slot specifies the duration of one superframe cycle. In addition, for convenience, the time slots within each frame cycle may be numbered consecutively. As a specific example, the network manager 27 may fix the duration of a slot at 10 milliseconds, and may define a superframe of length 100 to produce a 1 second frame cycle (ie, 10 milliseconds times 100). In a 0-based numbering scheme, the example superframe may include slots numbered 0, 1, . . . 99.

As discussed in more detail below, network manager 27 reduces latency and additionally optimizes data transmission by including multiple concurrent superframes of different sizes in network scheduler 67 . Also, some or all superframes of the network schedule 67 may span multiple channels or carrier frequencies. Accordingly, the main network schedule 67 may specify an association between each time slot of each superframe and one of the available channels.

Thus, master network schedule 67 may correspond to a collection of independent device schedules. For example, a network device such as valve positioner 34 may have a separate device schedule 67A. The device schedule 67A may only include information related to the corresponding network device 34 . Similarly, router 60 may have a separate device scheduler 67B. Accordingly, network device 34 may transmit and receive data according to device schedule 67A without knowledge of the schedules of other network devices, such as schedule 69B of that device 60 . For this purpose, the network manager 27 may manage the overall network schedule 67 and the respective individual device schedules 67 (eg, 67A and 67B), and communicate the individual device schedules 67 to corresponding devices when needed. Of course, the device schedules 67A and 67B are derived from the overall or main network schedule 67, and as a subset thereof. In other embodiments, individual network devices 25 and 35 - 50 may at least partially define or negotiate device schedules 67 and report these schedules to network manager 27 . According to this embodiment, the network manager 27 may assemble the network schedule 67 from the received device schedule 67 while checking for resource contention and resolving potential conflicts.

In order to optimally utilize the available wireless resources and ensure efficient and reliable data delivery, the network manager 27 may further optimize scheduling decisions in view of routing, alternatively optimize routing in view of scheduling decisions. In some particularly useful embodiments, the network manager 27 can conceptually combine routing concepts such as edges of a directed graph with scheduling resources such as time slots and superframes to define links. These links may further include several distinct types, such as dedicated links associated with a known pair of devices, shared links where at least one of the transmitter or receiver is not pre-assigned, Broadcast and multicast links addressed by multiple devices, etc. In these embodiments, network manager 27 may define a superframe by analyzing the topology of network 14, forming a directed graph of communication paths between a specified set of pairs of network devices, based in part on update rates at those network devices, and Sequentially assign time slots within defined superframes to directed connections between these devices to define a set of links specifying the direction and time of each data transmission to efficiently utilize link resources. Additionally, in those embodiments in which network 14 operates over several wireless channels, each link may specify the channel on which a particular transmission occurs. As discussed in detail below, network manager 27 can thus ensure that these network devices communicate efficiently and reliably. For example, network manager 27 may ensure that on a multi-hop communication path, a data packet spends as little time as possible before being transmitted to the next hop on the path.

The communication protocol supporting wireless network 14 described generally above is referred to herein as WirelessHART protocol 70 , and the operation of this protocol is discussed in more detail with reference to FIG. 2 . As will be appreciated, each of the direct wireless connections 65 can transmit data according to the physical and logical requirements of the WirelessHART protocol 70 . At the same time, the WirelessHART protocol 70 can efficiently support communication within time slots and on carrier frequencies associated with superframes as defined by the device-specific schedule 69 .

Figure 2 schematically illustrates the layers of an exemplary embodiment of the WirelessHART protocol 70, approximately aligned with the layers of the well-known ISO/OSI 7-layer model of communication protocols. For comparison, FIG. 2 additionally shows the layers of the existing "wired" HART protocol 72 . It will be appreciated that the WirelessHART protocol 70 does not necessarily have to have a wired counterpart. However, as will be discussed in detail below, the WirelessHART protocol 70 can greatly facilitate its implementation by sharing one or more upper layers of the protocol stack with existing protocols. As indicated above, as indicated above, the WirelessHART protocol 70 may provide the same or a higher degree of reliability and security than the wired protocol 72 serving similar networks. At the same time, by eliminating the need to install wires, the WirelessHART protocol 70 can provide several important advantages, such as reducing the costs associated with installing network equipment. It will also be appreciated that while FIG. 2 represents WirelessHART protocol 70 as a wireless counterpart to HART protocol 72, this specific correspondence is provided here as an example only. In other possible embodiments, one or more layers of the WirelessHART protocol 70 may correspond to other protocols, or as mentioned above, the WirelessHART protocol 70 may not even share the topmost layer with any of the existing protocols. application layer. It is also possible to use the WirelessHART protocol stack as a network layer according to other communication standards such as Foundation Fieldbus, Profinet, ModbusTCP, and Internet IP. In these cases, the WirelessHART protocol 70 may be responsible for transmitting real-time data, alarms, alarms, trends, or other information in accordance with the HART communication standard.

As shown in Figure 2, the wireless extension of HART technology can add at least one new physical layer (e.g., IEEE802.15.4 radio standard) and two data link layers (e.g., wired and wireless mesh) to the known Wired HART implementation. In general, the WirelessHART protocol 70 may be a secure, wireless mesh networking technology operating in the 2.4GHz ISM radio band (block 74). In one embodiment, the WirelessHART protocol 70 may utilize IEEE 802.15.4b compliant Direct Sequence Spread Spectrum Communication (DSSS) radios and channel hopping on a transaction-by-transaction basis. The WirelessHART communication may be mediated using TDMA to schedule link activity (block 76). Also, preferably, all communications are performed within designated time slots. One or more source devices and one or more destination devices may be scheduled to communicate in a given time slot, and each time slot may be dedicated to communication from a single source device, or the source devices may be scheduled to utilize Shared communication access mode like CSMA/CA for communication. A source device may send a message to one or more specific target devices or may broadcast a message to all target devices for which a timeslot is assigned.

Because the WirelessHART protocol 70 described here allows for the deployment of mesh topologies, an important network layer 78 can also be specified. Specifically, the network layer 78 may enable direct wireless connections 65 between individual devices and enable routing of data between a particular node of the wireless network 14 (e.g., device 34) and the gateway 22 through one or more intermediate hops. . In some embodiments, a pair of network devices 30-50 may establish a communication path comprising one or several hops, while in other embodiments all data may travel either upstream to gateway device 22 or downstream from gateway device 22. Propagate to a specific node.

To enhance reliability, WirelessHART protocol 70 may combine TDMA with a method of associating multiple radio frequencies with a single communication source (eg, channel hopping). Channel hopping provides frequency diversity that minimizes interference and reduces the effects of multipath fading. Specifically, data link 76 may generate an association between a single superframe and multiple carrier frequencies that data link layer 76 cycles between in a controlled and predefined manner. For example, the frequency bands available for a particular application of WirelessHART network 14 may have carrier frequencies F ₁ , F ₂ , . . . F _n . The associated frame R of a superframe S may be scheduled to occur with frequency F ₁ in cycle C _n , with frequency F ₅ in subsequent cycle C _n+1 , and with frequency F ₂ in cycle C _n+2 ,etc. The network manager 27 can use this information to configure relevant network devices, so that the network devices communicating in the superframe S can adjust the sending frequency or receiving frequency according to the current cycle of the superframe S.

The data link layer 76 of the WirelessHART protocol 70 may provide an additional channel blacklisting feature that restricts these network devices from using certain channels in the radio frequency band. The network manager 27 may blacklist radio channels in response to detecting excessive interference or other problems on the channel. Further, an operator or network administrator may blacklist channels in order to protect wireless services using a fixed portion of the radio frequency band that would otherwise be shared with the WirelessHART network 14 . In some embodiments, the WirelessHART protocol 70 controls blacklisting on a superframe basis, so that each superframe has an independent blacklist of prohibited channels.

In one embodiment, network manager 27 is responsible for allocating, assigning and adjusting time slot resources associated with data link layer 76 . If a single instance of network manager 27 supports multiple WirelessHART networks 14 , network manager 27 may generate an overall schedule for each instance of WirelessHART network 14 . The schedule may be organized into superframes containing slots numbered relative to the start of the superframe. In addition, the network manager 27 may maintain a global absolute slot count, which may reflect the total number of slots scheduled since the WirelessHART network 14 was started. This absolute slot count can be used for synchronization purposes.

WirelessHART protocol 70 may further define links or link objects for logically unified scheduling and routing. Specifically, a link can be associated with a specific network device, specific superframe, relative slot number, one or more link options (send, receive, shared), and link type (normal, discovery, broadcast, join) couplet. As shown in FIG. 2, the data link layer 76 may be frequency agile. More specifically, the channel offset can be used to calculate the specific radio frequency used to perform the communication. Network manager 27 may define a set of links in view of the communication requirements at each network device. Each network device can then be configured with the defined set of links. The defined set of links can determine when the network device needs to wake up, and whether the network device should transmit, receive, or both transmit/receive immediately after waking up.

With continued reference to FIG. 2, the transport layer 80 of the WirelessHART protocol 70 allows efficient best-effort communication and reliable, end-to-end acknowledged communication. As will be appreciated by those skilled in the art, best effort communication allows devices to send packets without end-to-end acknowledgment and does not guarantee data order at the destination device. User Datagram Protocol (UDP) is a well-known example of this communication strategy. In the process control industry, the method can be useful for publishing process data. Specifically, because devices periodically propagate process data, end-to-end acknowledgments and retries already limit utility, especially given that new data is produced on a regular basis. In contrast, reliable communication allows devices to send acknowledgment packets. In addition to ensuring data delivery, the transport layer 80 can also schedule packets sent between network devices. This approach may be preferred for request/response traffic, or when event notifications are sent. When using the reliable mode of the transport layer 80, communications can become synchronous.

A reliable transaction can be modeled as a master sending out a request packet and one or more slaves replying with a response packet. For example, a master device can generate a certain request and can broadcast the request to the entire network. In some embodiments, network manager 27 may use a reliable broadcast to tell every network device in WirelessHART network 14 to activate a new superframe. Alternatively, a field device such as sensor 30 may generate a packet and propagate the request to another field device, such as portable HART communicator 55 . As another example, an alarm or event generated by field device 34 may be sent as a request directed to gateway device 22 . In response to successfully receiving the request, gateway device 22 may generate and send a response packet to device 34 to acknowledge receipt of the alarm or event notification.

Referring again to FIG. 2, the session layer 82 may provide session-based communication between network devices. Sessions can be used to manage end-to-end communication at this network layer. A network device may have more than one session defined for a given peer network device. Almost all network devices can have at least two established sessions with the network manager 27 if desired: one for pairwise communications and one for network broadcast communications from the network manager 27. Additionally, all network devices may have one or several gateway session keys. These sessions can be distinguished by the network device addresses assigned to them. Each network device can track security information (encryption keys, ephemeral counters) and transmission information (reliable transmission sequence numbers, retry counters, etc.) for each session that the device participates in.

Finally, the WirelessHART protocol 70 and the WiredHART protocol 72 can support a common HART application layer 84 . The application layer of the WirelessHART protocol 70 may additionally include a sublayer 86 that supports automatic segmented transmission of large data sets. By sharing the application layer 84, the protocols 70 and 72 allow common encapsulation of HART commands and data and eliminate the need for protocol translation in the uppermost layers of the protocol stack.

dispatch communication

3-6 provide a more detailed illustration of the channel and time slot resource allocation supported by the data link layer 76 and network layer 78 of the WirelessHART protocol 70 . As discussed above with reference to FIG. 1 , the network manager 27 may manage the definition of one or more superframes and may associate individual time slots within each of the defined superframes with available channels (e.g., carrier frequency) associated with one of them. As a specific example, Figure 3 shows an available communication scheme within individual time slots, while Figure 4 shows an exemplary data exchange between several devices using the time slots of a certain superframe. Next, Figure 5 shows a possible association between exemplary time slots and several available channels, and Figure 6 is a schematic diagram of several concurrent superframes comprising the time slots shown in Figures 3-5.

Referring specifically to FIG. 3, two or more network devices can exchange data in a time slot 100, which can be a dedicated time slot shared by a sending device and a receiving device, or can be a time slot with more than one transmitter and/or Or a shared time slot for one or more receivers. In either case, time slot 100 may have transmit schedule 102 and receive schedule 104 . In other words, one or more transmitting devices may communicate during time slot 100 according to transmit time slot schedule 102 and one or more receiving devices may communicate during time slot 100 according to receive time slot schedule 104 . Of course, slot schedules 102 and 104 are substantially exactly synchronized and start at the same relative time 106 . During the time slot 100, the transmitting network device transmits a predetermined amount of data on a communication channel such as defined by a specific carrier radio frequency. In some cases, the sending network device may also expect to receive a positive or negative acknowledgment within the same time slot 100 .

Thus, as shown in FIG. 3, the transmit slot schedule 102 may include a transmit segment 110 for transmitting output data, preceded by a pre-transmit segment 112, and may include a transmit segment for receiving pairs transmitted during segment 110. Receive segment 122 of acknowledgment of data. The sending segment 110 and the receiving segment 122 may be separated by a transition segment 116 during which the corresponding network device may adjust eg hardware settings. Also, as discussed below, reception schedule 104 may include sections for performing functions that supplement the functions implemented in sections 112-122.

In particular, the sending device may send out during segment 110 an entire packet or flow segment associated with the capacity of time slot 100 . As mentioned above, the network schedule 69 may include shared time slots that do not belong exclusively to the independent device schedule 67 of one of the network devices 25 and 30-55. For example, a shared slot may have a dedicated receiver, such as gateway 22, rather than a single dedicated transmitter. When necessary, one of the network devices 25-60 may send unscheduled information, such as a request for additional bandwidth, in the shared time slot. In these cases, a potentially transmitting device may check whether a shared slot is available by performing a Clear Channel Assessment (CCA) in the pre-transmit segment 112 . In particular, the sending network device may listen to signals propagating on the communication channel associated with time slot 100 during pre-transmission segment 112 to verify that no other network device is attempting to use time slot 100 .

At the receiving end of time slot 100 , the receiving device may receive the entire packet associated with time slot 100 within packet reception segment 114 . As shown in FIG. 3 , packet receive segment 114 may start at an earlier point in time than transmit segment 110 . If desired, packet receive segment 114 may be extended beyond transmit segment 110 (not shown) to allow for minor timing mismatches. Next, in a transition segment 116, the transmit slot schedule 102 requires the transmitting device to switch radio modes. Similarly, receive slot schedule 104 includes transition section 118 . However, segment 116 can be shorter than segment 118 because the sending device can start listening for acknowledgment data earlier to avoid missing the start of the acknowledgment.

Still further, the transmission schedule 102 may include an acknowledgment reception section 122 during which the transmitting device receives acknowledgments sent during an acknowledgment transmission section 124 associated with the reception schedule 104 . The acknowledgment receive segment 122 may start before the acknowledgment send segment 124 and, if desired, end later than the acknowledgment send segment 124 to reduce the loss of acknowledgments. The sending device may, upon receipt of a positive acknowledgment, delete packets sent during the send segment 110 from the associated send queue. On the other hand, if no acknowledgment arrives or the acknowledgment is negative, the sending device may attempt to resend the packet in the next scheduled dedicated time slot or in the next available shared time slot.

As shown schematically in FIG. 4 , the number of time slots 100 discussed above may be organized into a superframe 140 . Specifically, a superframe 140 may include a (typically) infinite series of superframe cycles 150-154, each cycle including a set of slots, shown in FIG. 4 as slot 142 relative to slot number 0 ( TS0), the time slot 144 (TS1) with relative time slot number 1, and the time slot 146 (TS2) with relative time slot number 2. Accordingly, the size of the superframe 140 of FIG. 4 is three slots. In other words, each of the time slots 142-146 of the superframe 140 repeats in time every two intermediate time slots. Thus, for a slot of 10 milliseconds, the interval between the end of a slot with a particular relative slot number and the start of the next slot with the same relative slot number is 20 milliseconds. Conceptually, time slots 142-146 may be further organized into superframe cycles 150-154. As shown in Figure 4, each superframe cycle corresponds to a new instance of the sequence of time slots 142-146.

Master network schedule 67 may associate transmission and reception occasions of some of the network devices participating in wireless network 14 with specific time slots of superframe 140 . Referring again to FIG. 4 , network segment 160 schematically illustrates a partial communication scheme implemented between network devices 34 , 60 and 36 of FIG. 1 . To simplify the illustration of superframe 140 , network devices 34 , 60 and 36 are additionally designed as nodes A, B, C, respectively, in FIG. 4 . Thus, according to FIG. 4, node A sends data to node B, which in turn sends data to node C. As noted above, each of nodes A-C includes a device schedule 67A-C that specifies time slots and channels (eg, radio carrier frequencies) for transmitting and receiving data at the respective device. Main network schedule 67 may include some of all data information stored in individual device schedules 67A-C. More specifically, network manager 27 may maintain master network schedule 67 as an aggregate of schedules associated with each of network devices 30-50, including device schedules 67A-C.

In this example, the duration of time slot 100 (FIG. 3) may be 10 milliseconds, and network device A may report data to device C every 30 milliseconds. Accordingly, the network manager 27 may specifically set the length of the superframe 140 at three time slots in view of the update rate of the network device A. Further, network device 27 may assign time slot 142 (TS0) with relative number 0 to network devices A and B, with device A acting as a transmitter and device B acting as a receiver. Network manager 27 may further assign the next available time slot 144 with relative time slot number 1 (TS1) to be associated with the transmission from device B to device C. Meanwhile, slot 146 remains unassigned. In this manner, superframe 140 provides a scheme by which network manager 27 may allocate resources in network segment 160 to transfer data from device A to sent to device C.

In the example shown in FIG. 4, a network device at node A may store information related to timeslot 142 as part of its device schedule 67A. Similarly, a network device at Node B may store information related to time slots 142 (receive) and 144 (transmit) as part of its device schedule 69B. Finally, network device C may store information related to timeslot 144 in device schedule 67C. In at least some of these embodiments, network manager 27 stores information about the entire superframe 140, including an indication that time slot 146 is available.

Importantly, superframe 140 need not be limited to a single radio frequency or other single communication channel. In other words, the individual time slots 142-146 defining the superframe 140 may be associated with different radio frequencies on a permanent or floating basis. Also, the frequencies used by various devices need not always be contiguous in the electromagnetic spectrum. In one embodiment, for example, time slot 142 of each of superframe cycles 150-154 may be associated with carrier frequency _F1 , while time slot 142 of each of superframe cycles 150-154 may be associated with carrier frequency F1. Slot 144 may be associated with carrier frequency _F2 , frequencies _F1 and _F2 may or may not be contiguous in the electromagnetic spectrum.

In another embodiment, at least some of the time slots 142-146 may move around the assigned frequency band in a predefined manner. FIG. 5 shows an exemplary association between time slot 144 of FIG. 4 and channels 172-179 in available frequency band 170 (corresponding to frequency subbands F ₁ -F ₅ ). In particular, each of the channels 172-179 may correspond to one of the center frequencies _Fi , _F2 , ... _F5 , which center frequencies are preferably at the same offset from their respective neighbors. Preferably, the channels 172-179 form a contiguous frequency band covering the entire available frequency band 170, although in all embodiments the channels 172-179 need to be contiguous or form contiguous frequency bands. Superframe 140 may use at least a portion of frequency band 170 such that one or more of time slots 142-146 are scheduled on different carrier frequencies in at least two contiguous cycles.

As shown in FIG. 5 , during frame cycle 150, channel 176 (frequency F ₃ ) may be used by time slot 144, during frame cycle 152, channel 174 (frequency F ₄ ) may be used, and during frame cycle 154, Channel 178 (frequency _F2 ) may be used. Slot 144 may then "return" to channel 176 in the next superframe cycle 150A similar to cycle 150 . Each of the specific associations of timeslot 144 with one of channels 172-179 is shown as timeslot/channel tuple 144A-C. For example, tuple 144A designates scheduled time slot 2 in round 150 on channel 176 associated with center frequency _F3 . Similarly, tuple 144B designates scheduled time slot 2 in round 152 on channel 174 associated with center frequency _F4 . Meanwhile, channel 172 associated with center frequency _F5 may not be assigned to slot 2 during any of cycles 150-152. However, different time slots of superframe 140, such as time slot 146, may be associated with channel 172 during one or more of cycles 150-152.

In this example, the frequency assignment associated with superframe cycle 150 may repeat immediately after cycle 154 (as shown in FIG. 5 as cycle 150A), and, after two cycles of superframe 140, time slot 144 may Again corresponds to tuple 144A. In this manner, time slot 144 may periodically cycle through channels 176, 174, and 178. It will be appreciated that time slots 144 may similarly cycle through a greater or lesser number of channels, regardless of the length of superframe 140, provided, of course, that sufficient channels are available in frequency band 170. The association between a single time slot and multiple channels during the different superframe cycles discussed above with respect to FIG. 5 and referred to as "channel hopping" greatly increases the reliability of the wireless network 14 . Specifically, Channel Hopping Reduction Channel hopping reduces the probability that a pair of devices scheduled to communicate in a certain time slot of a superframe will not be able to send and receive data when a certain channel is congested or unavailable. Thus, for example, a failure of channel 174 prevents a device using slot 144 from communicating during frame cycle 152 , but does not prevent it from communicating during frame cycle 150 or 154 .

Referring again to FIG. 4 , device schedules 67B and 67C may include information about each of tuples 144A-C discussed above with reference to FIG. 5 . In particular, each of device schedules 67B and 67C may store an assignment of time slots 144 to one of channels 172-179 within each of cycles 150-152. The master network schedule 67 (FIG. 1) may similarly include this information. At the same time, device schedule 67A need not necessarily include information related to time slot 144 because the corresponding node A (device 34 ) does not communicate during time slot 144 of superframe 140 . In operation, devices 60 and 36 corresponding to Nodes B and C may prepare for data transmission and reception at the beginning of each time slot 144, respectively. To determine whether slot 144 currently corresponds to tuple 144A, 144B, or 144C, devices 60 and 36 may apply a global absolute slot count to determine whether slot 144 is currently in frame cycle 150 , 152 , or 154 .

In defining the network schedule 69, the network manager 27 may define multiple concurrent superframes in view of the update rates of the network devices 25 and 35-50. As shown in FIG. 6 , network schedule 69 may include superframe 140 of length 3 and superframes 190 and 192 . Superframe 190 may be a five-slot superframe and superframe 192 may be a four-slot superframe, although different superframes may have different numbers of slots and various superframes may have the same number of slots. As shown in Figure 6, these superframes do not necessarily have to be aligned for relative slot numbers. Specifically, at a particular time 194, superframe 190 may schedule a time slot with relative number 2 (TS2), while superframes 140 and 192 may schedule time slots with relative number 1 (TS1). Preferably, superframes 140, 190 and 192 are time synchronized so that within each of these superframes, each transition to a new time slot occurs simultaneously.

Each of superframes 140, 190, and 192 may be primarily associated with, or belong to, an individual network device 30-50 or a subgroup of network devices 30-50. subgroup. For example, superframe 140 shown in FIG. 4 may belong to node A (i.e., network device 34), and the length of superframe 140 may advantageously be chosen so that during each of cycles 150-154, node A is at Measurement data is sent to Node B during time slot 142 (TS0). If the wireless network 14 defines 10 millisecond time slots, node A sends data to node B every 30 seconds. However, if node A is reconfigured to report measurements every 50 milliseconds, network manager 27 alone, or in conjunction with node A, may reconfigure frame 140 to have a length of five slots. In other words, the length of each superframe may reflect the specific transmission requirements of a particular network device 30-50.

Alternatively, more than one network device 30-50 may use superframes for sending and receiving data. Referring again to FIG. 4 , while superframe 140 may be primarily associated with node A, node B (network device 60 ) may also periodically send data to node C (network device 36 ) in time slot 144 of superframe 140 . Thus, different time slots of a particular superframe can be used by different network devices to originate, route or receive data. In a sense, the time slots of each superframe can be understood as resources allocated to different devices, and specific priorities are assigned to the devices that own the superframe. Furthermore, it will be understood that each network device may participate in multiple superframes. For example, network device 34 in FIG. 4 may route data on behalf of other network devices (eg, network device 32 shown in FIG. 1 ) in addition to routing its own data by routing device 60 . Preferably, devices participating in multiple superframes do not schedule simultaneous communications in different superframes. Although only three superframes are shown in FIG. 6 , the wireless network 14 of FIG. 1 may include any number of superframes, each of these different superframes being based on may be of any desired or useful length depending on the type and frequency of communications performed in the

The methods described above with respect to Figures 1-6 can be applied to process control systems, such as those in which various devices report measurements or other data according to independent device schedules and during occasional, often unpredictable "bursts" of data process control system.

routing technology

As discussed above with reference to FIG. 1, it is important to consider the location of network devices 30-50 so that wireless network 14 can establish itself in an efficient and reliable manner. In some cases, it may be necessary to add a router 60 where factory equipment may block or severely impact wireless connectivity or where network devices are located far from each other. Accordingly, in this or similar circumstances, it is desirable for wireless network 14 to be "self-healing," ie, capable of automatically handling at least some of delivery failures. To meet this or other design requirements, wireless network 14 may define redundant paths and schedules so that in response to detecting a failure of one or more direct wireless connections 65, network 14 may route data via alternate routes. Also, these paths can be added and deleted without shutting down or restarting the wireless network 14 . Because some of these obstacles or sources of interference in many industrial environments may be temporary or movable, wireless network 14 may be able to automatically reorganize itself. More specifically, in response to one or more predetermined conditions, pairs of network devices (e.g., 32 and 34, 60 and 36, etc.) may recognize each other as neighbors, thereby creating a new direct wireless connection 65 or vice versa , cancel the existing wireless connection 65 . Additionally, network manager 27 may create, delete, or temporarily suspend paths between non-neighboring devices.

Whether a particular network configuration is permanent or temporary, wireless networks 14 generally require fast and reliable methods of routing data between nodes. In one available embodiment, network manager 27 may analyze information about the layout of the network, the transmission capabilities and update rates of each network device 32-50, and other relevant information. The network manager 27 can then define routes and schedules in view of these factors. In defining routing and scheduling tables, network manager 27 may identify wireless network 140 as conforming to one of several network topologies compatible with the routing and scheduling techniques of the present disclosure.

Figures 7-9 schematically illustrate some of these network topologies. For clarity, each of Figures 7-9 shows a bidirectional connection between a pair of devices. However, it will be appreciated that each of the topologies shown in FIGS. 7-9 are also compatible with unidirectional connections or mixed bidirectional and unidirectional connections (ie, including bidirectional and unidirectional connections). Furthermore, each connection shown in Figures 7-9 may support several unidirectional connections in one or both directions, eg, each unidirectional connection is associated with a particular transmission instant. Referring specifically to FIG. 7, the network 350 may have a star network topology. Star network 350 includes routing device 352 and one or more end devices 254 . Routing device 352 may be a network device arranged to route data, while end device 254 may be a network device arranged to send data only for itself and only receive (or decode) data addressed to end device 254 . Of course, routing device 352 may also be a recipient or originator of data and may perform routing functions, among other tasks. As shown in FIG. 7 , end device 254 may have a direct connection 265 to routing device 352 , but end device 254 cannot be directly connected in a star topology. The direct connection 265 may be a direct wireless connection 65 or a wired connection.

Terminal device 254 may be the same type of physical device as routing device 352 and may be physically capable of routing data. The routing capabilities of the terminal 254 can be disabled during installation of the terminal 254 or during operation of a corresponding network (eg WirelessHART network 14 ). Furthermore, the routing capabilities of the terminal device 254 may be disabled by the terminal device 254 itself or by a dedicated service such as the network manager 27 . In a sense, star network 350 corresponds to the simplest topology available. It may be suitable for small applications requiring low power consumption and low latency. Additionally, it will be noted that star network 350 is deterministic, so there is only one route available between routing device 352 and a particular end device 254 .

Referring now to FIG. 8, network 270 is arranged in a mesh network topology. Each network device of mesh network 270 is a routing device 352 . A mesh network provides a robust network with multiple paths between various devices. In wireless applications, mesh networks are best able to adapt to changing radio environments. For example, device 274 of network 270 may transmit data to device 276 via intermediate hop 278 or intermediate hop 380 as long as the corresponding path 382-388 allows transmission in that direction. As shown in FIG. 8 , paths 382 and 384 enable routing device 274 to send data to routing device 276 , providing redundancy and resulting increased reliability to network 270 .

Another type of network topology is shown in FIG. 9 . Network 390 combines elements of star and mesh topologies. Specifically, the star mesh network 390 includes a number of routing devices 352 (labeled "R") and end devices 254 (labeled "E"). Routing devices 352 may be connected in a mesh fashion and may support redundant paths. The selection of a particular topology may be performed automatically by a network element such as the network manager 27, or by a user configuring the network. Specifically, the user may choose to override the topology selected by the network manager 27 or the default topology associated with the WirelessHART protocol 70 . It is assumed that in most applications, the mesh near topology may be the default topology because of the topology's inherent reliability, availability, and redundancy. Clearly, since a WirelessHART device can act as a router device, several different configurations are compatible with the same physical configuration of field devices and routers.

Source routing and graph routing can be applied to the topologies discussed with reference to Figures 7-9. While both types of routing can be equally useful in different situations, graph routing will be discussed first. In general, in mathematical theory and applications, a graph is a set of vectors (nodes such as 352 and 254) and edges (direct connections 65 and 265). For example, WirelessHART protocol 70 or another protocol serving wireless network 14 or 140 may use a graph to configure a path connecting a communication endpoint such as device 30 to gateway 22 shown in FIG. 1 . In some embodiments, the graph and associated paths are configured by network manager 27 . Network manager 27 may also configure individual network devices, such as field devices 30 - 40 , router 60 , etc., with partial maps and routing information, which may be stored in connection table 69 . Wireless network 14 may contain multiple graphs, some of which may overlap. Further, a certain network device may have paths through the device in multiple graphs, some of which may direct data to the same neighbor of the device. Preferably, each graph in the network is associated with a unique graph identifier.

The protocols serving wireless networks 14 and 140, such as WirelessHART protocol 70, can be configured to operate with many different topologies to support various application requirements. As a result, wireless network 14 or 140 may concurrently support several routing methods such as unidirectional graph routing and source routing. While existing examples of wireless networks support both approaches, it will be appreciated that wireless network 14 or 140 may additionally support bidirectional graph routing or may route data using only one of these techniques. However, regardless of the type and number of concurrent routing techniques, each device on wireless network 14 or 140 may be assigned a unique network address. Once some form of unambiguous identification about other network elements is known to every potential receiver of the data, it may be done by a stand-alone device such as the field devices 30-40, by a centralized dedicated service such as the network manager 27, or by This centralized service works with individual devices to make routing-related decisions. As indicated above, at least one possible implementation of wireless network 14 may rely on network manager 27 to implement most or all routing decisions and communicate related data to network devices 30 - 50 for storage in connection table 69 . Further, routing decisions can be made at the point of origin (ie, at the source of the data packet) or at a central location. Also, routing decisions can be adjusted at each intermediate stop or "hop" in the path of a packet from a source point to a destination. Well-known optimization techniques, such as Dijkstra's classical algorithm for finding the least-cost path to a particular node, can be used to optimize graphs.

In the examples discussed below, wireless networks provide at least two methods of routing, which can be based on the specific requirements and conditions of a given system, such as the physical layout of the network elements that make up the system, the number of elements, the or the desired amount of data to send from each element, etc., to select. Moreover, both methods may be used by the wireless network simultaneously, and each method may, in view of certain aspects of the performance of each of the two methods, be selectively applied to particular types of data or Applies to a specific host or host group. For example, a measurement of a process variable or a command to open a valve may tolerate relatively little delay in delivery, and wireless network 14 may accordingly employ the faster and more reliable of the two methods. Meanwhile, device configuration commands or responses may tolerate longer delays and may be suitable for other methods.

As briefly indicated above, for certain distributed control networks, especially for networks connecting devices in the process control industry, it is common to direct data to certain devices for management, diagnostics, log collection, and other purposes. 10-12 illustrate several perspectives of a wireless network 300 that enables data transmission in two general directions: toward gateway 202 (referred to herein as the "upstream" direction) and away from gateway 202 (referred to herein as the "upstream" direction). "downstream" direction). For security reasons, network 300 does not allow direct data transfer between peer field devices, although the techniques described herein can be used in such cases if desired.

FIG. 10 shows upstream routing in network 300 . Specifically, the network manager 302A (or backup network manager 302B) may define several directed graphs, each graph including the network access point 305A or the second network access point 305B as terminal nodes. In other words, each graph's path in exemplary network 300 leads to and terminates at one of two network access points 305A or 305B. Specifically, diagram 310 (indicated by bold solid arrows) may include network devices 312, 314, 316, 318 and gateway 305A, wherein those paths associated with diagram 310 may include direct wireless connections 320, 322, 324, 326 and 328. Diagram 340 (indicated by bold dashed arrows) may include network devices 312 , 316 , 318 , 342 and gateway 305A, and paths including direct wireless connections 344 , 346 , 348 , 350 and 352 . In directed graph 310 , network device 312 may be referred to as the head of directed graph 310 and network access point 305A may be referred to as the tail of directed graph 310 . Similarly, network device 312 is the head of directed graph 340 and network access point 350 is the tail of directed graph 340 . Network manager 302A, or under certain operating conditions alternate network manager 302B, may define maps 310 and 340 and may communicate complete or partial definitions of these maps 310 and 340 to network devices 312-318 and 342. As discussed above with reference to FIG. 1, network devices 312-318 and 342 may maintain an up-to-date version of connection table 69 that stores these partial path definitions. In some embodiments, the network access points 305A-B may not require information about the graphs 310 and 340 if the respective communication path terminates at one of the network access points 305A-B. However, it will be appreciated that the network access points 305A-B may also originate data and may store information related to one or more graphs having paths originating outside the network access points 305A-B. It will be further noted that in general the path of a certain graph may pass through the network access point 305A or 305B as an intermediate node. However, the exemplary network 300 defines paths that always either originate from a network access point 305A or 305B or terminate at a gateway device 305A or 305B.

To send a data packet along a certain graph, the source network device may include the identifier of the graph in the packet header or trailer in the data packet. The packet can traverse those paths corresponding to the graph identifier until it either reaches its destination or is discarded. For example, in order to be able to route packets in the graph 310, the connection table 69 of each network device belonging to the graph 310 may contain entries including the graph identifier and the addresses of neighboring network devices (1) belonging to the same , and (2) one hop closer to the destination. For example, network device 316 may store the following connection tables:

graph identifier node Figure_310 318 Figure_340 318 Figure_340 342

The network device 342 can store the following information in the connection table:

graph identifier node Figure_340 302

Although the above exemplary connection table merely lists devices associated with a particular entry, it will be noted that the node column of the connection table may store addresses of neighboring devices as defined by the addressing scheme of network 300 or WirelessHART network 14 .

In another embodiment, the node column may store the aliases of neighboring devices, an index to an array storing the full or short addresses of these neighbors, or any other means of unambiguously identifying network devices. Alternatively, the connection table may store map identifier/wireless connection tuples as shown below.

graph identifier connect Figure_310 326 Figure_340 346 Figure_340 348

In other words, the connection table may list one or more direct wireless connections 65 corresponding to a particular map. For example, network device 316 may query the connection table and send a packet carrying map identifier 340 via direct wireless connection 346 or 348 .

As shown in Figure 10 and in the table above, redundant paths can be established by associating more than one neighbor with the same graph identifier. Accordingly, data packets that arrive at network device 316 and include map identifier 340 in packet headers and trailers may be routed to network device 318 or to network device 342 . While performing routing operations, network device 316 may perform a lookup in the connection table by virtue of graph identifier 340 and send the packet to either (or both) network devices 318 or 342 . Also, routing between two or more possible hops can be random or performed according to a predefined algorithm. For example, this selection may be made considering load balancing purposes or in view of delivery statistics. Thus, network device 316 may know, either through a peer network device or from network manager 27, that when routing a packet along graph 340, selecting network device 318 as the next hop has a lower probability of successfully delivering the packet or has Longer expected or average latency. Network device 316 may then attempt to route more or possibly all packets associated with graph 340 to network device 342 .

In one embodiment, the neighbor device confirms the receipt of the data packet by sending an acknowledgment packet. In the above example, network device 316 may release the packet as soon as neighbor network device 318 or 342 acknowledges receipt of the packet. On the other hand, if the acknowledgment is not received within a predefined period of time, the network device 316 may attempt to resend the packet or route the packet via an alternate hop or path. Additionally, network device 316 may collect statistics on successful delivery attempts and failed delivery attempts. Subsequent routing decisions, eg, choosing between hops 318 and 342, may include or be based on the adjusted statistics. Of course, network device 316 may apply statistics relating to network devices 318 and 342 to other correlation graphs, and may also communicate these statistics to other network devices either directly or via network manager 27 .

As discussed above, in the graph routing approach, a network device sends a packet with a graph identifier in the network packet header along a set of paths to a destination. Importantly, the graph identifier alone is sufficient to route the packet, and each packet can be delivered correctly based on the graph identifier alone, although other routing information may also be included in the packet header. On the way to the destination (ie, on the path), all network devices can be pre-configured with graph information specifying the neighbors to which these packets can be forwarded. Because graph routing requires pre-configuration of intermediate network devices for each possible destination, graph routing can be better suited for communication from network devices to gateways and from gateways to network devices.

Referring now to FIG. 11 , network manager 302A or 302B may also support downstream routing with respect to one or both of gateways 305A-B. Specifically, graph 380 (indicated by solid bold arrows) may include nodes 315, 314, and 312, and direct wireless connections 382-386. Network access point 305A is the head of graph 380 and wireless device 312 is the tail of graph 380 . Meanwhile, map 390 (indicated by bold dashed arrows) may similarly connect network access point 305A to wireless device 312 with network access point 305A acting as the head of map 390 . However, graph 390 may include nodes 305A, 318, 342, 316, and 312, as well as direct connections 392-298. Thus, to send a data packet to wireless device 312, network access point 305A may include a map identifier corresponding to map 380 or 390 in a header or trailer of the data packet. It will be appreciated that each of diagrams 380 or 390 may also include dual connection paths to ensure reliability, and that in general, network manager 302A or 302B may use techniques similar to those discussed above with reference to FIG. 10 . Also, it will be noted that the connection table 69 for each of the wireless devices 312-318 and 342 may include graph routing information related to downstream graphs and upstream graphs for routing purposes.

Additionally, as shown in FIG. 12, wireless network 300 may use source routing. In source routing, there is no need to pre-configure the relay device. To send the packet to its destination using source routing, the source network device may include in the header of the packet, for example, an ordered list of devices through which the packet must travel. An ordered list of devices can effectively define the communication path for that packet. When the packet passes through the specified path, each routing device can extract the next node address from the packet to determine where the packet should be propagated next, that is, the next packet should be sent to in the next hop where. Therefore, source routing requires the topology of the network 14 to be known in advance. However, if a certain network device does not find itself on the routing list, the network device may send the packet back to the first device specified in the source routing list. Source routing allows packets to go to arbitrary destinations without intermediate devices being established explicitly or in a predetermined configuration.

For example, network device 312 may send the packet to network access point 305A by specifying the full path in the packet header or packet trailer. As shown in FIG. 9, network device 312 may generate a list 310 containing the addresses of network devices 314, 315, and 305A and send this list 310 along with the packet to the first hop or device on the list, the network device 314. Network device 314 may then traverse list 310 , locate the identity of network device 314 , extract the field from list 310 , identify network device 315 as the next hop to receive the packet, and finally send the packet to network device 315 . The source routing list can reside in an optional field of the network packet header and can be of variable size depending on the number of hops to the destination. Similarly, network device 315 may traverse table 310, locate its own address and identity, and send the packet to the next hop or device in list 310 (in this case, network access point 305A).

In general, only those network devices that have obtained complete network information from the network manager 27 or 302A-B use source routing because only the network manager 27 or 302A-B knows the complete topology of the network. An additional limitation of source routing is that it does not provide redundancy in intermediate network devices, since each packet is originated with a header or trailer that explicitly specifies each intermediate hop and does not provide any routing alternatives. Therefore, if one of the intermediate network devices specified by the header or trailer fails to relay the packet, delivery of the packet fails. Therefore, it is the responsibility of the network manager 27 or 302A-B to detect the failure and reprogram or reconfigure the source with an alternate route. To facilitate detection of these error conditions, the wireless network device 14, 140 or 300 may require the network device to send a routing failure notification to the network manager 27 or 302A-B. Accordingly, a protocol such as WirelessHART protocol 70 may provide a message type or information element in the protocol definition to report this or other type of delivery failure. In another embodiment, route list 310 (see FIG. 9 ) may specify alternate routes in addition to the route selected by the sender. In yet another embodiment, the primary route and one or more backup routes may be partially merged to avoid repeating common parts of the path in the packet header and trailer.

Referring generally to Figures 1, 3 and 10-12, the network manager 27 or 302A-B may maintain a list of all devices in the network. The network manager 27 or 302A-B may also contain the entire network topology, including the complete map of the network and the latest part of the map that has been communicated to each device. The network manager 27 or 302A-B may use the information the network manager receives from the network devices 30-40, 50, 60, 55, etc. to generate routing and connection information. The network manager 27 or 302A-B may then build a map of the network from the list of network devices and the neighbors reported by each network device. Referring again to FIG. 1 , for example, network device 50B may report “seeing” neighbor devices 60 and 34 . The network manager may also be responsible for generating and maintaining all routing information for the network. In one embodiment, there is always one full network route and several purpose-specific routes that are used to send setpoints and other settings from gateway 302A or 302B to the final control command (FIGS. 7-9). Additionally, a broadcast route (which flows through most or all devices in the network) can be used to send broadcast messages from network manager 27, 114 or 302A-B to all devices in network 14 or 300. Furthermore, once the routing information and the burst mode update rate are known, the network manager 27, 114 or 302A-B can also implement the scheduling of network resources.

When a device is initially added to a network 14, 140 or 300, the corresponding network manager may store all neighbor entries reported from each network device. The network managers 27, 114 or 302A-B can use this information to build initial complete network maps and modify these maps during runtime. The network map is combined, optimizing several attributes including: hop count, reporting rate, power usage, and overall traffic flow as reflected by the statistics collection discussed above. A key aspect of this topology is the connection list that connects devices together. Because the existence and health of individual connections may change over time, the network manager 27, 114 or 302A-B may be additionally programmed or configured to update the overall topology, which may include adding and deleting information in each network device. In some embodiments, only the network manager 27, 114 or 302A-B and the gateway 22 or 302A-B may know enough information to use source routing. More specifically, it may be desirable to prevent peer-to-peer communication between any two arbitrary devices for security purposes. Furthermore, although graph routing and source routing have been described here as occurring between a field device and a gateway device, these types of routing can be used between any two devices in the network, including, for example, any two field devices or Between network devices, between any two gateway devices, etc.

In short, graph routing can direct traffic upstream and downstream for the network manager 27 or gateway 22, and both graph routing and source routing can be optimized for low-latency applications, including traffic from network devices to the Measurement information from gateways and control information transmitted from gateway devices to final control commands such as regulating valves, on-off valves, pumps, fans, baffles, and motors used in many other ways.

In some embodiments, path redundancy may be a matter of policy of the network manager 27, 114 or 302A-B rather than a coincidental overlap of graphs. In other words, networks 27, 114 or 302A-B may attempt to define at least two neighbors for each device. Accordingly, network managers 27, 114 or 302A-B may be configured to actively pursue either a mesh topology or a star mesh topology. Therefore, supporting protocols such as the WirelessHART protocol 70 can provide high end-to-end data reliability. From a physical standpoint, each field device or other network device should be within communication range of at least two other devices that can receive messages from that field device and forward those messages.

The network manager 27, 114 or 302A-B may additionally verify each graph definition to ensure that no loops have been formed. In embodiments where the network managers 27, 114 or 302A-B actively pursue path redundancy and define many graphs of various sizes, communication paths may sometimes be incorrectly defined to direct packets from a source back to the same source. Depending on such an error graph definition, packets may be routed from the source directly back to the source or one or more intermediate hops may be visited in between back to the source. Loop verification may be performed whenever the topology of the associated network changes, for example due to the addition or removal of devices, or whenever the network manager 27 adjusts the routing maps and schedules for any reason. Alternatively, network manager 27 may periodically perform loop checking as a background task.

Combining Routing and Scheduling Decisions

In a wireless network such as wireless network 14 or 300, the same graph routing can be used with several schedules. In particular, packets may be sent along the same route while updating the network schedule changes and the moment the packet is sent from or to a node. In this sense, routing and scheduling can be conceptually and functionally separated to facilitate network management. On the other hand, however, network manager 27 may perform routing and scheduling substantially in parallel to achieve robustness and improve wireless network 14 or 300 performance and reliability. More specifically, network manager 27 may make at least some scheduling decisions in view of relevant routing constraints, and conversely make routing decisions in view of scheduling constraints. In some particularly useful embodiments, the network manager 27 may first analyze the topology of the network 14 or 300, build a network map 67, and then proceed to view the network map 67 and specific device parameters such as transmission rate, power capacity, etc. To build the network scheduling table 69.

When making scheduling decisions in view of the network topology, the network manager 27 can conceptually associate slots in a particular superframe with edges of a particular directed graph (which are direct connections 65 in the example discussed here) to Define a convenient combined space and time unit, the link. Specifically, a link may be associated with a direct connection 65 between two communication endpoints and the times at which the two communication endpoints exchange data.

Further, the network manager 27 may associate slots with several types corresponding to different principles for allocating and using slots. Specifically, a specific pair of network devices 32-50 can share a dedicated unicast link, so that one network device of the pair of network devices can use a designated time slot to send information to the other network device of the pair of network devices. Of course, as discussed above with respect to FIG. 3, a network device that transmits certain information during a time slot may also receive a corresponding acknowledgment from the device that received the information, and in this sense, each of the network device pair Network devices act as transmitters and receivers during a single time slot. However, for simplicity, a device sending a block of information in a slot is referred to herein as a "talker", and a device receiving the block of information is correspondingly referred to as a "listener".

In contrast to dedicated links, shared links can have more than one talker, but only one listener. In a sense, the shared link is still a unicast link, since the link has only one listener. Broadcast and multicast links, on the other hand, can have one talker and many listeners. Further, a dedicated link has one talker and a restricted set of listeners.

According to another aspect, a particular network device may consider a dedicated, shared, directed or broadcast link as either a transmit link or a receive link. Referring again to FIG. 10, for example, wireless devices 312 and 314 may share a dedicated link associated with a particular time slot and direct wireless connection 330 so that wireless device 312 transmits data over the link and wireless device 314 transmits data over the link. Receive data. Accordingly, wireless device 312 may view the link as a transmitting link, while wireless device 314 may view the same link as a receiving link. Therefore, splitting the link into transmit and receive links is a matter of device perspective. It will also be noted that certain links may be defined or used as Send/receive link.

As indicated above, each link may be associated with a certain time slot, regardless of the type of link. In operation of wireless network 14 or 300, network manager 27 or 302A-B may assign links to network devices 30-50, 312, 314, 316, 318, and so on. Depending on the link type, a network device associated with the link may send packets, receive packets, or remain idle. If the packet's destination matches one or more neighbors on the other end of the link, a network device operating with the sending link or the sending/receiving link can send package. On the other hand, a network device having a receive link or a transmit/receive link on which no packets are transmitted listens for one or more incoming packets during the time slot associated with that link. In those embodiments that also use a shared link, preferably the device performs Clear Channel Assessment (CCA) or another method of preventing resource contention before the device begins transmitting over the shared link. In at least some embodiments, all devices participating in dedicated or shared links must wake up and listen during the time slots associated with those dedicated or shared links.

As explained above with reference to Figure 3, during a single time slot, one network device may send a data packet and the other device may reply with an acknowledgment. Similarly, a communication session over a link may include the transmission of data packets and the transmission of acknowledgments, which may be affirmative ("ACK") or negative ("NACK"). In general, a positive acknowledgment may indicate that the receiver has successfully received the packet and has assumed ownership of the packet for further routing if the receiver is not the final destination of the packet. At the same time, a negative acknowledgment may indicate that the receiver cannot receive the data packet at this time but has detected that the packet is error-free. Further, both ACK and NACK can carry timing information so that corresponding devices can correctly maintain network synchronization. In some embodiments, packets sent to a unicast network device address may require a link layer acknowledgment within the same time slot, while packets sent to a broadcast network device address (eg, 0xFFFF) may not require an acknowledgment.

For example, when network manager 27 defines a direct connection 65 between network devices 30-50, these network devices receive link assignments. These devices may accordingly define and maintain a corresponding device schedule 67 (see Figures 1 and 4). A link assignment may in part specify how the network device should use a certain time slot in a superframe. Thus, each link may consist of exactly one time slot, type assignment (i.e. transmit and/or receive), neighbor information or other data identifying edges of the bidirectional or unidirectional graph associated with that link, and its additional Send and/or receive properties.

In some embodiments, the device scheduler 67 of each network device 30-50 may maintain additional flags or indicators to properly maintain various types of links. For example, device scheduler 67 may set a shared flag for each shared link so that the corresponding network device 30-50 can properly access the link for transmission. Further, with respect to the shared link, the network 14 or 300 may use the well-known slotted Aloha contention management algorithm to define the lifetime of the shared link. Accordingly, network devices 30-50, 305A-B, 312, 314, 316, 318, etc. may employ a collision avoidance scheme with backoff (delay) in case of collision. In some embodiments, this delay may be implemented as a time measurement independent of the duration of a single slot. In other particularly useful embodiments, the backoff may be implemented with a delay measured in an integer number of time slots. Specifically, a device that has encountered a collision may back off by the duration of a single slot, two slots, etc. from the beginning of the next scheduled slot. By synchronizing the back-off interval with the time slot, the device can optimize the retry mechanism and ensure that retry attempts only occur when there is a possibility of transmission. Using a shared link may be desirable when the bandwidth requirements of the devices are low and/or traffic occurs sporadically or in bursts. In some cases, using a shared link can reduce latency because the network device does not need to wait for a dedicated link, although this is usually only true when the chance of collisions is relatively low.

Form an effective scheduling and routing scheme

In addition to optimizing routing by analyzing the network topology, the network manager 27 can define graphs and allocate resources during scheduling in view of the types of data that particular network devices can transmit and the expected transmission frequency of each type of data at each particular device . More specifically, WirelessHART protocol 70 can support several network communication services. Both the existing HART protocol 72 and the WirelessHART protocol 70 support exchanging request/response data, publication of process data, sending broadcast messages, and block data transfer of large data files. The WirelessHART protocol 70 can also support the transfer of management data such as network configuration data and device communications such as periodic measurements reported by field devices using the same protocol and the same resource pool, enabling more efficient scheduling and routing. high efficiency.

The network manager can allocate communication resources for each network device according to the amount of data that the network device can publish per unit time. For example, a WirelessHART flow meter 30 in the WirelessHART network 14 may have an update rate of four seconds, while a WirelessHART pressure sensor 32 may have an update rate of ten seconds. Operators can configure network devices 30 and 32 with these values according to the specific needs of the process control system in which WirelessHART network 14 is implemented. As indicated above, multiple superframes may be used to define different communication schedules for various network devices or groups of network devices. Initially, network manager 27 may reserve a specific superframe for all network manager requests. After accepting network devices such as flow meters 30 and pressure sensors 32, network manager 27 can allocate additional superframes for the four-second and ten-second communication rates and assign the additional superframes to the network Devices 30 and 32. Network manager 27 may also define superframes for alarms and network events, respectively, before or after network devices 30 and 32 are added to WirelessHART network 14 . Network devices 30 and 32 may, but are not required to, participate in one or more superframes concurrently. By configuring specific network devices to participate in multiple overlapping superframes of different sizes, multiple communication schedules and connectivity matrices can be established that can work concurrently without scheduling conflicts. In addition, because some critical applications, such as asset management and specific device applications, often require considerable bandwidth for short durations, the network manager 27 may also generate additional temporary superframes as needed. For example, a user may issue a request to view or change a device's configuration or to generate a diagnostic screen. The network manager 27 may support this temporary increase in demand for communication slots by defining additional superframes with additional slots that may remain valid for several minutes (this is an example only).

The network manager 27 may consider both the update rate of the network devices and the topology of the WirelessHART network 27 when creating the directed graph. However, the network manager 27 can also make graph routing decisions in a scheduling-independent manner. For example, the network manager can add, delete or update maps while leaving the network schedule intact. More specifically, the network schedule may have time slots available in defined superframes that the network manager 27 may use as resources when defining new maps or updating existing maps. In this sense, the WirelessHART protocol 70 allows map configuration decisions to be made independently of scheduling decisions. This feature of the WirelessHART protocol 70 can allow the WirelessHART network 14 to respond more quickly to changes in the environment and the operating state of the network devices, because the WirelessHART protocol 70 can quickly and in a non-trivial manner by changing only a portion of the existing configuration. Intrusive way to adjust routing.

It will be appreciated that some of the methods discussed above are not necessarily limited to data packets, and are applicable to other communication technologies. For example, the network may use a circuit-switched method, and instead of traveling in packets of limited size, data may be transmitted as a stream over dedicated channels between communication terminals.

As simplified examples, Figures 13-20 and the accompanying figure description illustrate efficient routing schemes and corresponding formation of a communication schedule applying the concepts discussed above to a four-node mesh network. Specifically, FIG. 13 schematically illustrates a network 400, several direct wireless connections formed between nodes A, B, C, and D (or network devices 402-408) of the network 400, and in view of factors such as network topology, signal Factors of strength and/or quality, latency requirements, etc. may form edges 410-420 of several directed graphs in network 400. To further simplify this example, only upstream (ie, device-to-gateway) data propagation is discussed with reference to FIGS. 13-20. For clarity, the network nodes will be referred to below as nodes A-D, and the devices corresponding to these nodes will be referred to as gateway device 402 or field devices 404-408. As further shown in FIG. 13, nodes A-D may store, maintain and update device-specific connection tables 422-428. Next, Figures 14-16 illustrate the formation of several superframes to support data exchange between nodes A-D. Finally, FIGS. 17-20 illustrate example device-specific communication schedules 432-438 that may correspond to nodes A-D of FIG. 13 .

Referring specifically to FIG. 13 , the wireless network 400 may include a gateway device 402 that operates as a node A in the network 400 and connects the wireless network 400 to an external network (not shown) or a stand-alone external device (not shown). Due to typical business requirements, this gateway device may be a live device, ie connected by a cable or wire to a substantially unlimited power source. Nodes B-D, on the other hand, may be battery powered field devices. Of course, in other embodiments some or all of nodes B-D may also be connected to power lines or other energy sources. As also shown in FIG. 13 , each of the field devices 404 - 408 may have a specific update rate at which the device sends data to another node, such as the gateway device 402 . For example, field devices 404 and 406 may generate outgoing data every second, and field device 408 may generate outgoing data every four seconds.

In general, Nodes B-D may also correspond to any type of device used or equipped with communication capabilities. It is contemplated that, for example, Nodes B-D may be personal computers operating in a home network. However, since the routing and scheduling techniques therein are particularly useful in process control and sensor mesh networks, in this particular example Nodes B-D are operating in a process control environment and performing various sensing and measurement functions (e.g., temperature, pressure, , flow, acidity, valve actuator position, etc.) or field devices that control functions (starting, positioning, etc.). Field devices 404-408 may exchange measurement and control data over any protocol that supports the routing and scheduling techniques discussed above with reference to FIGS. 3-12. In one embodiment, the protocol supporting these functions may be the WirelessHART protocol 70 shown in FIG. 12 .

Network manager 440 may be a software module running within node A. Similar to network manager 27 shown in FIG. 1 or network managers 302A-302B shown in FIGS. 10-12 , network manager 440 may be responsible for making routing and scheduling decisions in network 440 . Specifically, network manager 440 may initially gather device and signaling information from each of nodes B-D, as well as information related to its own host node A, and define an initial network map for network 400 in view of these factors. More specifically, each of nodes A-D may report a Received Signal Strength Indication (RSSI) value specifying signal energy to each of its potential neighbors. Alternatively, nodes A-D may report measured signal quality or any other measurement that can be used to assess signal quality. In addition, each of nodes A-D may report to network 400 parameters such as power supply capacity (e.g., battery-powered, line-powered, etc.), projected or actual delivery demand (e.g., every second, every four seconds measured value updates, etc.; scheduled occasional automatic updates; updates only in response to requests; etc.) and other information related to the operation of the field devices in the network 400 .

After collecting device information from nodes A-D, network manager 440 can establish the topology of network 400 by selecting direct wireless connections between pairs of neighboring devices to define one or more directed graphs between pairs of nodes A-D. In the particular example shown in Figure 13, each directed graph has either node A as the head or tail. Network manager 440 may thus enable data flow upstream and downstream in the network relative to gateway device 42 .

When defining a direct wireless connection, the network devices 402-404, or if desired, the network manager 440, may compare the strength or quality of the wireless signal transmitted by one of the nodes A-D or measured by another of the nodes A-D with a Thresholds are compared to determine if the signal is sufficient to support a direct wireless connection. In at least some embodiments, each of the network devices 402-404 automatically determines with which of the remaining network devices 402-402 the network device can establish a direct connection, thereby making the corresponding device its neighbor. To this end, each of network devices 402-404 may execute the same routine for collecting and analyzing signal measurements. Referring to FIG. 13, for example, network device 408 may determine that the signal emitted by the transceiver of gateway device 402 cannot provide a strong enough direct wireless connection between these devices, in other words, between nodes A and D. direct wireless connection. On the other hand, because device 404 is closer to device 402, there are fewer obstacles between devices 402 and 404, or other factors, network device 404 at node B can measure the same signal from node A to obtain a more accurate signal. Good (eg greater intensity or higher quality) measurements. Network device 404 may accordingly determine that a potential direct wireless connection exists between nodes A and B. Preferably, but not necessarily, network device 402 makes a similar determination and also decides that a potential direct wireless connection exists between nodes A and B.

Next, each of network devices 402 - 408 may report the collected measurements and potential direct connections to network manager 440 . Alternatively, network 400 may be formed iteratively such that node A initially forms a complete and operational single-node network, nodes B and C later join network 400 to form a complete and operational three-node network, and node D finally joins network 400 As a neighbor of nodes B and C to form a complete and operational four-node network. In some cases, and particularly in large networks with a large number of network devices, a particular node may establish direct connections with many potential neighbors. According to some embodiments, network manager 400 may direct network devices to disable some of these redundant direct connections, eg, based on the relative quality of these direct connections.

In the formed network 400 shown in Figure 13, nodes B and C may each have a direct wireless connection to node A (supporting directed connections 410 and 412, respectively). In other words, each of field devices 404 and 406 may be one hop away from gateway device 402 . Additionally, Nodes B and C may also share a direct wireless connection to support a directed B-to-C connection 414 and a directed C-to-B connection 416 . Meanwhile, node D may be connected to node A only through node B, only through node C, or through both nodes B and C. Thus, node D may be two hops (D to B to A) or three hops (D to B to C to A) away from node A.

To achieve the definition of directed connections 410-420 and further the directed routing graph of network 400, network manager 440 may consider several factors in addition to the set of available direct wireless connections between neighboring network devices. Specifically, the network manager 440 may attempt to define a directed graph starting at nodes B-D and ending at node A using the smallest number of hops possible. As a result, the directed graph from node B to node A includes a single directed connection 410 . Similarly, the directed graph from node C to node A includes a single directed connection 412 . A directed graph from node D to node A requires at least a single intermediate node B or C, but may also include two intermediate nodes. Applying the principle of minimum hop count described above, network manager 440 may define a directed graph that connects D to A that includes, for example, a sequence of connections 418 and 410 . It will be noted that the directed graph connecting A to D need not include the same intermediate hop B (as discussed in more detail above with reference to Figures 10-12).

Further, the network manager 440 may prefer to use the electrified device as an intermediate node. If, for example, node B is battery powered and node C has an unrestricted power source, then network manager 440 defines the graph as going from node D to node A via node C, all else being equal. However, in the example shown in Figure 13, only node A is connected to the unrestricted power supply.

In some embodiments, network manager 440 may also compare the relative signal strengths of available direct connections reported by these network devices to select between similar paths. Assuming, for example, that node D reports a stronger signal from node B than from node C, the network manager 440 may prefer a directional path from node D to node A that includes node B as an intermediate hop.

Continuing to refer to FIG. 13 , network 400 may thus include the following directed graph definition for upstream (ie, device-to-gateway) data propagation:

identifier source destination definition G ₁ B A B->A _G2 C A C->A G ₃ D A D->B->A G ₄ D A D->C->A

Each of the device-specific connection tables 422-428 may store a corresponding portion of these graph definitions in accordance with the principles discussed above.

Network manager 440 may apply additional principles to define an efficient directed routing graph for network 400 . In some embodiments, network manager 440 may limit the number of hops in the graph to three or fewer. Of course, other constraints may also be employed. In some embodiments, the network manager 400 may also impose a limit on the number of neighbors of a network device. If, for example, a network device is able to detect signals sent by ten neighbor devices, the network manager 440 can pare down that number to allow the network device to establish direct connections with only three or four network devices, the three or four network devices Devices are selected based on signal strength or some other criterion.

After defining the directed graph between some of the nodes AD, the network manager 440 may proceed to define a network schedule based on the graph decisions made above. In this example, the network manager 440 may enforce the following additional restrictions and/or principles: 1) limit the number of concurrent active communication channels to 16; 2) not allow any of the network devices 402-408 to be in the same time slot Listening twice; 3) Allowing more than one device to send to the same target device (for example, allowing broadcast links and dedicated links to coexist in the same network schedule); 4) Scheduling early hops on multi-hop paths to shorter Before the late jump; and 5) support an update rate conforming to the formula update rate = 2 ^x , where x is a positive or negative integer value (i.e. enabling 1 second, 2 seconds, 4 seconds, 8 seconds, 16 seconds seconds, 32 seconds, etc. update rate selection). It should also be noted that in the examples discussed with reference to FIGS. 13-20 , network manager 440 knows the "raw" network topology (i.e., all measurements of signals from potential neighbors reported by each network device), network devices 402-408 Each of the network devices 402-408 stores and maintains connection tables 422-428, and the network manager 440 knows the data update rate of each of the network devices 402-408. As an additional requirement, network 400 may implement a redundancy scheme whereby each transmission is configured with one retry opportunity on one path and one retry opportunity on the other path. Of course, other update rates, routing constraints, redundancy schemes, etc. are possible, and it should be appreciated that the principles listed above are provided as examples only.

Generally, the network manager 440 can use the following strategies when defining a superframe: 1) set the length of the superframe according to the data update rate of the device associated with the superframe; 3) starting from the non-remote device farthest from the gateway 402, assign a link to each intermediate network device in the path to the gateway 402; 4) schedule a retry opportunity for each transmission on the main link , where possible, schedule another retry opportunity on the redundant link; 5) schedule each network device to receive at most once in a time slot; and 6) assign time slots at a faster The update rate starts.

In addition to device-specific superframes, network manager 440 may define network management superframes to propagate network management over network 400 and to receive service requests from network devices 402-408. Specifically, the network manager 440 may set the length of the network management superframe equal to the longest device superframe. In some embodiments, it may be desirable to further enforce a lower limit on the length of the administrative superframe, such as 600 slots (ie, one minute in the ten millisecond slot embodiment). Further, network manager 440 may reserve a set of time slots for advertising, so that potential network devices wishing to join network 400 may send join request messages during these advertising time slots. Next, the network manager 440 traverses the network graph according to a breadth-first search starting from the gateway device 400, and numbers each encountered device as N ₀ , N _{1 .} . . N _n .

As shown in FIG. 14, the network manager 440 may define a superframe 450 of 100 slots for one-second updates of Nodes B and C. It should be noted that in this exemplary embodiment, network manager 440 creates a single superframe for both nodes with the same update rate. However, in other embodiments, each node may be associated with a separate superframe. To improve reliability of transmission, superframe 450 may run over several channels (as discussed with respect to FIG. 5, for example), shown in column 452 as offsets 0-3. Also, each of columns 454 designates a particular slot within superframe 450 of 100 slots. Each cell in the table showing superframe 450 designates one or more talkers and one or more listeners for each offset/slot tuple, respectively.

Continuing with FIG. 14, the network manager 440 may begin assigning time slots for the directed graph ending at node A by starting at node B, since node B has the fastest update rate of one second. (Of course, in this example, node C has the same update rate and is also located one hop away from node A. In this and similar cases, network manager 440 may apply any desired tie-breaking (tie- braking) technique to choose between two or more equivalent candidates). Since Node B has a direct connection 410 to Node A as part of the directed graph _G1 , the network manager 440 can assign to Node B time slots T0 and T1 on the offset 0 channel to periodically update the data sent to node A. Specifically, the Node B may use slot T0 for transmitting data in a scheduled manner and slot T1 for retries. FIG. 14 shows these channel-slot-directed connection assignments as dedicated links 460 and 462 . In view of the discussion provided above, it should also be appreciated that Node B sees dedicated links 460 and 462 as dedicated transmit links, while Node A sees these links as dedicated receive links.

It should be appreciated that although in this particular example time slots T0 and T1 are associated with the same channel, retry opportunities could be assigned to different channels (eg, offset 1) to diversify the utilization of the channel. Alternatively, the network manager 440 may instruct each of the nodes A-B to perform channel hopping according to a predetermined order and for a network schedule that does not directly specify frequency hopping. For example, each of nodes A-B may always shift the channel offset by 1 every superframe period, thereby ensuring that the relative order specified by the network schedule remains unchanged.

Next, the network manager 440 may allocate time slots TS3 and TS4 on the 1 offset channel to the directed connection 412 from node C to node A. Similar to links 460 and 462, dedicated links 464 and 466 may correspond to the same scheduled update in a redundant manner. Additionally, as part of diversifying channel utilization, links 464 and 466 are assigned to different channels than links 460 and 462 in this example.

Continuing with the example, the network manager 440 may then define a superframe 470 of 400 slots for Node D's four second update. Since superframes 450 and 470 may start at the same time, time slots TS0 and TS1 on the 0 offset channel may be considered occupied and therefore unavailable. Also, since node D can transmit its update data to node A via node B along graph _G3 , slots T0 and T1 shares may not be allocated to directed connection 418 at all, regardless of channel offset. This limitation can be easily understood by observing that during time slots T0 and T1 Node B is already engaged in communication, ie sending data on links 460 and 462 . For this reason, the earliest time slots that network manager 440 can allocate to directed connection 418 are TS2 and TS3. Of course, the 1-offset channel is already occupied by communication nodes A and C in time slots TS2 and TS3, so the directed connection 418 can reserve the still available channel with an offset of 0 (links 472 and 474 in FIG. 14 ).

As a next step, network manager 440 may schedule directed connections 420 associated with the secondary path from node D to node A (shown above in Figure _G4 ). Dedicated link 476 may accordingly receive an assignment to time slot TS4 on the 0 offset channel. In the same time slot TS4, link 478 associated with directed connection 410 may reserve an adjacent channel (offset 1). It should be noted that Node B may use link 478 to forward to Node A data received by Node B from Node D in time slot TS2 or TS3. In other words, link 478 may be associated with graph _G3 . Applying similar principles, network manager 440 may then define links 480-484, also shown in FIG. 14 .

Referring now to FIG. 15 , the network manager 440 can also define a management superframe 500 of 400 slots based on the length of the longest data superframe 470 . Unlike superframes 450 and 470, management superframe 500 may include several shared links 502-508 reserved for advertising, several dedicated Links 510-516, and several dedicated links 520-526 reserved for propagating join responses from network manager 440 residing at node A to terminal nodes B-D of network 400.

Referring specifically to shared links 502-508, it should be noted that each of nodes A-D expects to receive data from any potential device other than one of the existing nodes of network 400 over these shared channels. Thus, in at least some embodiments, links 502-508 provide an opportunity for external devices to submit requests to join network 400 and eventually become new network nodes. At the same time, links 510-516 serve the respective paths for conveying these requests from terminal nodes B-D to node A.

FIG. 16 shows other links that the network manager 440 may reserve in a management superframe 500 . More specifically, links 530-536 may transmit command data from nodes B-D to network manager 440 at node A. Links 540-546 may accordingly support Node A's transmission of command responses in the opposite (ie, downstream) direction.

Additionally, network manager 440 may assign each of superframes 450, 470, and 500 a unique superframe identifier. To simplify maintenance, debugging, and visualization, the network manager 440 may assign the identifier "0" to a management superframe, the identifier "1" to a 1-second superframe, and the identifier "4" to a 4 second superframe. In this way, the frame identifier may conveniently convey at least some information about the frame to an engineer or operator.

For superframes 450, 470, and 500, in general, it should be appreciated that the particular manner in which network manager 440 assigns time slots within these superframes prevents overlapping collisions in frame cycles following the first cycle. Specifically, superframe 450 goes through four frame cycles during a single cycle of superframe 470 or superframe 500 . Thus, time slot TS0 of superframe 450 would coincide with time slots TS0, TS100, TS200, and TS300 of superframes 450 and 500 if all superframes initially started at the same time. The method of allocating time slots described above ensures collision-free overlap of multiple superframes such as 400, 450, and 500 by conforming participating network devices to an update rate of ^2X that provides optimal divisibility of superframe lengths.

Therefore, in the manner described above, the network manager 440 can generate several directed graphs G1-G4 to efficiently route data in the network 400, define several data superframes 450 and 470 for data update, define at least one management superframe Frames 500 to propagate advertisement information, join requests and responses, commands and command responses _{, and} other types of network management data, and then schedule dedicated, shared and possibly other types of links. It should be appreciated that the allocation of time slots shown above minimizes the amount of time a transient data packet takes between reaching an intermediate node in a communication path and departing for the next node. For example, a packet leaving node D via node B in the direction of node A at time slot TS2 may proceed from node B to node A at time slot TS4 at the earliest.

To enable each node A-D to function according to these definitions, the network manager 440 may then assign map and schedule definitions to each node A-D. 17-20 illustrate several example device-specific schedules 432-438 consistent with those shown in FIGS. 14-16 (see FIG. 13).

Referring to FIG. 17 , Node A specific device schedule 432 may specify, for each entry, frame identifier 602 , slot number 604 , channel offset 606 , peer identifier 608 , link option 610 , and link type 612 . In other embodiments, the schedule 432 may include alternative or additional information such as MAC addresses of neighbor devices. In this particular example, each row of schedule table 432 may specifically indicate an action, if any, performed by device 402 in a particular time slot. For ease of readability, FIG. 17 lists the entries of the schedule table 432 in ascending order of time slots. However, each of the nodes A-D can use any suitable method, such as an array, a linked list, a binary tree, etc., to store the corresponding schedule.

For example, according to entry 620 of schedule table 432, node A may know that at time slot TS4, node A must switch its transceiver to receive mode, tune to a channel with offset 1, and expect to receive from Node B receives data. Furthermore, node A may know that the transmission is associated with superframe 4 (ie, a 400-slot superframe). As another example, entry 626 may correspond to slot 8 and may specify a relationship to a management superframe. Since node A is scheduled to receive advertisements from candidate network devices, node A may not know the sender of the data (marked with an asterisk).

It should be noted that the example schedule 432 only shows the allocation of time slots 0-21. However, superframe 450 may have time slots TS0, TS1, ... TS99, and superframes 470 and 500 may have time slots TS0, TS1, ... TS399. Although FIGS. 17-20 do not show unassigned time slots, it should be appreciated that, for example, node A neither receives nor transmits data during the duration of time slots TS22 through TS99 of the shorter superframe 450 .

18-20 illustrate respective device-specific schedules 434, 436, and 438. Referring to FIG. As can be seen from these figures, each of schedule tables 432-438 specifies only schedule information related to a particular node A-D. Similarly, each of connection tables 422-428 may store relatively little information that still specifies all routing actions related to the corresponding device. Thus, Node B does not need to know routing or scheduling information at Node C, for example. This approach can significantly reduce the amount of management traffic transmitted over the network, since the network manager 440 does not need to share the entire topology formed and the complete network schedule with each device. In addition, network 400 can be made more secure because none of nodes A-D can know where or when the rest of network 400 is transmitting information (with the possible exception of node A, which in at least this example acts as network manager 440 host).

Although the description of scheduling and routing performed in the wireless network described here has been described in detail for a four-node wireless network for simplicity, the principles described here are applicable to any size, with any number of nodes, graphs, etc. network of. Likewise, these figures can be of any desired or required size or length.

While the foregoing has described in detail a number of different embodiments, it should be understood that the scope of this patent is defined by the words of the claims presented at the conclusion of this patent. The detailed description should be construed as exemplary only, without describing every possible embodiment, since describing every possible embodiment would be either impossible or practical. Numerous alternative embodiments could be implemented, using either current technology or technology developed after the filing date of this patent, which would still fall within the scope of the claims.

Claims (31)

1. transmit the method for data in the communication network of an operation under process control environment reliably, this method comprises:

Based on the topology of this communication network, generate first routing diagram with a plurality of nodes and one group of edge, comprising:

Carry out related with the corresponding network equipment in a plurality of network equipments of participating in this communication network each node in said a plurality of nodes; And

Carry out the corresponding connection between two network equipments in each edge in this group edge and the said a plurality of network equipments of participating in this communication network related; And

At least based on this first routing diagram, define the communication scheduling table of this communication network, comprising:

Be each edge assigned communication time slot in this group edge of this first figure according to predefined procedure, said communication time slot is included in transmission and reception opportunity of the above a plurality of network equipment of communication channel of said communication network.

2. the method for claim 1 wherein generates first routing diagram and comprises: generates the directed graph with source node and destination node, make data use this directed graph along propagating towards unique direction of this destination node from this source node.

3. method as claimed in claim 2 is wherein carried out related comprise with communication time slot with intramarginal each edge of this group of this first figure according to predetermined order: first scheme from the order assignment said communication time slot of this source node towards this destination node propagation with this.

4. method as claimed in claim 2 wherein generates first routing diagram and further comprises: based on the available power supply type in each node place in one group of both candidate nodes, from this group both candidate nodes, select node of graph; And each node that wherein should organize in the both candidate nodes can be as the intermediate node between this source node and this destination node.

5. method as claimed in claim 2 further comprises: based on the topology of this communication network, generate the second oriented routing diagram with a plurality of nodes and one group of edge; Wherein in source node or the destination node of in the source node of this second directed graph or the destination node and this first figure corresponding one identical; And the communication scheduling table that wherein defines this communication network is at least based on this first figure and this second figure.

6. the method for claim 1, wherein this communication network is a wireless network; And related comprising carried out in the corresponding connection that wherein will organize between two network equipments in each edge and the said a plurality of network equipments in the edge: based on the signal quality that joins with this join dependency, from one group of candidate connects, select to connect.

7. the method for claim 1, the communication scheduling table that wherein defines this communication network further comprises:

Use the communication time slot of predefined single hop duration; And

Transmission requirement based at least one node in the node that is included among this first figure defines first superframe; Wherein this first superframe comprises the cyclic sequence of repetition, and each circulation comprises the communication time slot that the quilt of first number is dispatched continuously, the length of these these first superframes of communication time slots definition; Each communication time slot of wherein assigning for the edge in this group edge is associated with this first superframe.

8. method as claimed in claim 7, the communication scheduling table that wherein defines this communication network further comprises:

Definition comprises second superframe of repetitive cycling sequence, and each circulation comprises the communication time slot that the quilt of second number is dispatched continuously, the length of these these second superframes of communication time slots definition; Wherein this second number is not equal to this first number.

9. method as claimed in claim 8 wherein defines the transmission requirement of second superframe based on another node at least in the node that is included among this first figure; Wherein this method further comprises:

Topology based on this communication network generates secondary route figure, and this secondary route figure has a plurality of nodes different with this first routing diagram; Wherein defining the communication scheduling table comprises: under the situation of length less than the length of this second superframe of this first superframe, assign with communication time slot that this second superframe is associated before, the communication time slot that appointment is associated with this first superframe.

10. method as claimed in claim 7, the communication scheduling table that wherein defines this communication network further comprises: carry out related with network management data this first superframe.

11. the method for claim 1; The communication scheduling table that wherein defines this communication network comprises a plurality of superframes of definition; Each superframe in said a plurality of superframe comprises the repetitive cycling sequence of the communication time slot that the quilt of some is dispatched continuously, wherein this number ability enough 2 ^xExpress, x is an integer.

12. the method for claim 1, the communication scheduling table that wherein defines this communication network further comprises: for assigning main communication time slot and secondary communication time slot in each edge; The transmission network equipment that wherein is associated with particular edge sends data to the reception network equipment that is associated with this particular edge during this main communication time slot, send the result to generate; And send the result based on this and during this pair communication time slot, send data to this reception network equipment conditionally.

13. a generation is used for the method for the communication scheduling table of mesh communication network, this mesh communication network has a plurality of nodes and operation under process control environment, and this method comprises:

Definition is the communication time slot of duration fixedly;

With repetitive cycling sequence definition first superframe, each circulation comprises the communication time slot that the quilt of first number is dispatched continuously, the length of these these first superframes of communication time slots definition; And

The node that the time slot allocation that will be associated with this first superframe is given this communication network between direct the connection, wherein time slot be assigned to each connect with this communication network of appointment in one group of routing diagram of one or more routed path in a routing diagram be associated.

14. method as claimed in claim 13 further comprises:

Each direct connection that a channel allocation in many channels is assigned to time slot.

15. method as claimed in claim 14, each the bar channel in wherein said many channels is the radio bands that is associated with corresponding centre frequency.

16. method as claimed in claim 13, wherein each figure in this group routing diagram is the directed graph of the way flow of the data of definition from the source node to the destination node.

17. method as claimed in claim 16; Wherein in the source node of each directed graph or the destination node is the field apparatus of implementation controlled function under process control environment, and in this source node or this destination node another is the gateway device that this mesh communication network is connected to external network.

18. method as claimed in claim 17 wherein defines first superframe and comprises: according to the length of this first superframe of rate selection of process of transmitting control data between this field apparatus and this gateway device.

19. method as claimed in claim 13 further comprises:

With repetitive cycling sequence definition second superframe, each circulation comprises the communication time slot that the quilt of second number is dispatched continuously, and these communication time slots define the length of second superframe; Wherein this second superframe is associated with the network management data of this communication network; And each node in wherein said a plurality of node transmits and receive data at least one time slot of this second superframe.

20. method as claimed in claim 19, wherein this communication network is a wireless network; And wherein the corresponding multicast conversation on the wireless channel at least some time slots in the time slot of this second superframe and a plurality of wireless channels is associated.

21. method as claimed in claim 19, wherein this communication network is a wireless network; And wherein the single receiving node at least one time slot in the time slot of this second superframe and the said a plurality of nodes and one or more nodes of not participating in the potential node of this communication network are associated.

22. method as claimed in claim 13 further comprises:

Equipment particular schedule information is distributed at least some nodes in said a plurality of node, comprises:

For each of another node in said a plurality of nodes directly connects assigned timeslot, many communications

One of one of channel and transmission or receiving mode.

23. a raising comprises the method for reliability of the based on wireless mesh network of a plurality of nodes, this method comprises:

Set up a plurality of direct connections, each in wherein said a plurality of direct connections directly is connected to the uni-directional wireless with sending node and receiving node and is connected;

Based on the topological generation definition node of this wireless network between a plurality of directed graphs of communication path, wherein each directed graph comprises at least one in said a plurality of direct connection;

With the quilt of predefined single hop duration continuously the repetitive cycling of the communication time slot of scheduling define a plurality of concurrent superframes, the length of this superframe of number of time slot definition in each the concurrent superframe in wherein said a plurality of concurrent superframes; And

Define a plurality of primary links to generate the communication scheduling table of this based on wireless mesh network, comprising:

With one in each bar primary link and the said a plurality of direct connections directly connect carry out related, this directly connect with said a plurality of directed graphs at least one be associated; And

For each primary link distribute with said a plurality of superframes in a single time slot that superframe is associated.

24. method as claimed in claim 23 wherein generates a plurality of directed graphs and comprises:

First node in said a plurality of nodes is appointed as the head node of this based on wireless mesh network;

Go at least one inbound figure of data flow of the inbound direction of this head node along the direct-connected first subclass generation definition; And

Leave at least one departures figure of data flow of the outbound direction of this head node along the direct-connected second subclass generation definition.

25. method as claimed in claim 23; Wherein at least some nodes in a plurality of nodes of this based on wireless mesh network are execution control function or measurement function and announce the field apparatus of report data with specific scan speed in Process Control System, and wherein define a plurality of concurrent superframes and comprise:

Define first superframe that has with the time slot of corresponding first number of first length according to the sweep speed of primary scene equipment;

Define second superframe that has with the time slot of corresponding second number of second length according to the sweep speed of secondary scene equipment; Wherein

This first length is less than this second length.

26. method as claimed in claim 25; Wherein, each primary link comprises: first time slot allocation is given be used to send first primary link from the report data of this primary scene equipment for distributing the time slot in one of said a plurality of superframes; Then second time slot allocation is given and be used to send second primary link from the report data of this secondary scene equipment; Wherein this first primary link joins with the direct join dependency of initiating from this primary scene equipment, and this second primary link joins with the direct join dependency of initiating from this secondary scene equipment.

27. method as claimed in claim 23 further comprises:

A plurality of carrier radio frequencies that distribution is used by this wireless network, each carrier radio frequency in wherein said a plurality of carrier radio frequencies is identified through unique channel offset; And

Wherein defining many primary links further comprises to generate the communication scheduling table: carry out related with one of said channel offset each bar link.

28. method as claimed in claim 23 further comprises:

Define many secondary links as a part that generates said communication scheduling table; Be associated with primary link and secondary link comprising at least one the direct connection in the said a plurality of direct connections at least one directed graph in said a plurality of directed graphs, and wherein different time slots is assigned to this primary link and this pair link.

29. method as claimed in claim 28, wherein continuous slot is assigned to this primary link and this pair link.

30. method as claimed in claim 28; Wherein generating one group of directed graph comprises: defined node between the dual communication path that comprises main communication path and secondary communication path, wherein this main communication path and this pair communication path are different at least one directly is connected.

31. method as claimed in claim 23; Wherein setting up one group directly connects and comprises: according to the intensity of radio signal detected by the first node in two nodes and that sent by the Section Point in these two nodes, between these two nodes, set up uni-directional wireless and connect.

Images